SLOPSHOPPER

negative-space

When an edit adds authorization, input validation or error handling, lists the refusal, invalid-input, failure and boundary-value tests that no test touched…

newbandguardcommand
★ 1v0.1.2MITupdated 2026-10-16simota/claude-mods/mods/negative-space
A shopper browsing a rack in a slop shop
Preview · a replayed session in a sandbox
claude · ~/work/app · negative-space
› fix the failing auth test and add an audit log call ⏺ Read(src/auth.ts) ⎿ Read 6 lines ⏺ Update(src/auth.ts) ⎿ Added 2 lines, removed 1 line ⏺ Bash(bun test) ⎿ 3 pass, 1 fail ● Done. refresh now rejects expired claims and logs an audit event. ✻ Worked for 42s · done 4:20 PM › /negative-space ⎿ negative-space: 🕳 このセッションの編集に、認可・入力検証・エラー処理の追加は見つかっていません。 ────────────────────────────────────────────────────────────────────────────────────────────────────────────────────── › ? for shortcuts
README

claude-mods

English | 日本語

Personal Claude Code mods (function-hook plugins), published as one marketplace.

ModWhat it does
agent-paneLive pane of the session's subagents: running ones with elapsed time, finished ones with result and duration
done-gateFlags unverified edits, stops repeated failing commands, and blocks git --no-verify
sandbox-lensTallies sandbox denials seen in Bash output and reports them with /sandbox-report
skill-hintSuggests up to three matching skills above the prompt from what was typed; never invokes them.
todo-agentCollects #TODO(agent) follow-ups from Claude's replies into a per-project list with a /todos pane.
context-meterBand above the prompt with a Gruvbox sparkline of context use per turn, plus tokens and cost
tool-rowRedraws finished tool-call rows in the terminal as one compact line with a Nerd Font icon, Gruvbox colors and a short stat
image-previewThumbnails of the images a session writes, reads or generates: under the tool's row in the terminal, and in an /images pane
self-auditAfter each turn, a cached fork of the conversation audits Claude's own reply against the CLAUDE.md reporting rules and flags violations above the prompt
context-dietTrims the skill listing to what the project kind needs (off by default); /context-diet shows what was dropped and the tokens saved
browser-camLive pane of the browser Claude drives: the latest screenshot from Claude in Chrome, the page URL and the recent actions
prompt-coachHolds back a typed prompt that looks ambiguous (no done-condition, unclear target) and offers guesses at what it meant, sending it as is, or adding one; on by default
loopbackFeeds what self-audit, done-gate and sandbox-lens found back to Claude beside the next prompt, instead of only showing it to the person, skipping rules hit-rate found mostly wrong; off by default
zukanError book: catches the errors Bash commands hit across sessions and repos, records what got past each one, and when one comes back tells Claude and the person how it was fixed last time
pr-lampThe current branch's pull request as a lamp: checks, review, unresolved threads and what is left before a merge, in the status line and a /pr pane; red checks and new review comments go to Claude beside the next prompt
fog-mapExploration map: the repository as a tree under fog, cleared where Claude read and marked where it edited, made a file or hit an error, with the importers an edit may have broken unseen; /map
assertion-guardCatches edits that weaken tests to pass: assertions removed, expectations loosened or rewritten, tests skipped, errors swallowed; asks Claude for the reason beside the edit
negative-spaceWhen an edit adds authorization, validation or error handling, lists the refusal and boundary-value tests no test touched this session covers; /negative-space
mutation-tap/mutate runs a mutation test limited to the lines changed since HEAD and lists the mutants the tests did not catch
perf-delta/perf compares a benchmark before and after (a saved baseline, or a git ref in a temporary worktree): median time, peak memory, spread, and whether the change beats the noise
handoff-capsuleBefore a compaction and at session end, keeps the goal, decisions, open items, commands to reproduce and files touched; the next session picks which parts to bring back
port-lanternThe dev servers the session started: URL, port, whether each still listens, and their latest log lines, in the status line and /ports
artifact-shelfA shelf of what sessions made (CSV, PDF, HTML, video, patches…), tied to the call and request that made each; /shelf searches by kind, word and time
focus-hudOne status line: goal, target, what is running and for how long, and what the session waits on; /focus
side-effect-ledgerA ledger of what a git diff never shows (push, publish, deploy, PRs, API writes, messages) with outcome, how to undo, and whether the undo was checked; /ledger
trust-labelLabels external text (web, issues, logs, browser pages, services) with its source beside the result and flags candidate injected instructions; a heuristic, never a safety verdict; /trust
visual-diffCompares browser screenshots of the same URL and size before and after an edit: side by side, a map of changed cells, regions, and the edits in between; /vdiff
journey-traceRecords Claude's browser actions and the expected states as a replayable recipe; /journey md for a bug report, /journey e2e for a Playwright draft
hit-rateGives each finding of done-gate, assertion-guard, self-audit and zukan a later verdict from what the session did next, keeps each rule's precision across sessions, and names the rules loopback stops feeding; /hit-rate; off by default
token-lensChecks an element's actual colours, spacing and type against the project's design tokens and lists off-token values with the nearest token; /token-lens <selector>

Install

At the prompt of a terminal session:

/plugin install <mod> --marketplace simota/claude-mods

From a shell:

claude plugin install <mod> --marketplace simota/claude-mods

This repository is private: installing needs read access on GitHub and a git credential for github.com on the machine (an SSH key, or gh auth login then gh auth setup-git).

Usage

Options are set with /plugin configure <mod>@simota-mods in a session, or from a shell:

echo '{"enabled":"true"}' | claude plugin configure <mod>@simota-mods --values-stdin

A change made with /plugin configure reloads the mod; one made from a shell applies after a restart. Bands above the prompt stack: each mod draws its row above what the others draw.

agent-pane

  • /agent-pane opens the pane: running subagents first with elapsed time, then finished ones with ✓/✗ and duration; a background run that was stopped shows ■ 停止. A background run that left the roster without a result shows ? 結果不明; a teammate or remote launch shows ↗ 起動済み(追跡外).
  • In a fullscreen terminal the pane opens by itself once per session, on the first Agent call (a reload does not reopen it).
  • The status line shows agents: N running while any run.

done-gate

Works on its own, no command:

  • A turn that edited files without a passing test, lint, typecheck or build afterwards leaves a band: 未検証の変更があります. It stays until a check succeeds or 閉じる is pressed. Only the program a command runs counts (ls build, npm ci are not checks), and only when its exit status is the command's: the check is last, or only && follows it (npm test | tail does not count).
  • The second identical failure of the same Bash command tells Claude to stop retrying and diagnose the root cause (again after the command has succeeded in between).
  • Any git command with --no-verify (or a prefix of it from --no-ver), and git commit -n / -nm, is denied, also inside subshells, $(…), backticks, bash -c "…", eval and wrappers such as sudo/timeout. The value of an option (-m "…"), a comment or a heredoc never triggers it. So is a git command that moves its hooks for the run (-c core.hooksPath=…, --config-env core.hooksPath=…) or carries a hook manager's off switch in its environment (HUSKY=0, LEFTHOOK=0, pre-commit's SKIP=…), or moves the hooks through config in its environment (GIT_CONFIG_KEY_n=core.hooksPath, GIT_CONFIG_PARAMETERS). ANSI-C quoting ($'--no-\x76erify'), env -S "…", bash -c -- '…' and a shell reading its script from stdin (bash <<< '…', a heredoc, echo '…' | sh) are read through. It is a best-effort guard: git aliases (git ci -n), a hook path written into config beforehand and an environment exported by an earlier command are not detected.

sandbox-lens

  • /sandbox-report lists the sandbox denials seen in Bash output, for this session and all time: denied hosts and paths, counts, last seen, an example command, and which setting would allow each.
  • /sandbox-report reset clears the record.
  • Hints only: it never changes sandbox settings. A loopback or private address keeps its port in the hint (127.0.0.1:3000), since an entry without a port opens every port.
  • A <sandbox_violations> block, and Operation not permitted lines a program printed about itself, are read. For a command that succeeded or was interrupted, the block is taken from where the engine puts it, apart from the command's own output, so one inside a shown file or page is ignored even when it ends the output. A failed command keeps only its error text: the block that ends it (before an agent-proxy note) is read, and there a block the command printed itself cannot be told from the harness's. Credentials in the example command (tokens, passwords, Authorization/Cookie and other credential headers such as X-API-Key, URL userinfo, keys in env assignments including run-together names such as PGPASSWORD, known key shapes) are masked. Each section shows the top 10; when the record is full (200), the least recently seen entry gives way.

skill-hint

After a prompt is sent, up to three skills matching it show above the prompt (💡 skill候補: /coding-debug …), only ones that exist in the session. Nothing is invoked; 閉じる or the next prompt (a slash command included) clears it. A prompt that already names a skill (/name anywhere, or a hyphenated skill name such as coding-debug) gets no hint; built-in commands such as /clear and path segments such as src/review/ do not count as naming one. Plugin skills (plugin:name) are suggested by their full name.

todo-agent

  • Lines starting with #TODO(agent): <action> (after an optional quote or list bullet) in Claude's main-loop replies are collected per project (repo root or cwd). A mention inside a sentence, a table or inline code is not a TODO; subagent replies are ignored.
  • /todos opens the pane of this project's open items, each with a 完了 button.
  • /todos all lists every project's open items as text, marked as stored data.
  • The status line shows TODO(agent): N while the project has open items.

context-meter

A band with a sparkline of context use per turn and ctx 63% ▲4 · 126k · turn in 120k / out 8k · $1.23: the context size of the last request, then the whole turn's input/output (shown when the turn reported usage). The percentage turns yellow from 50% and red from 80%.

OptionDefault
bandtrueDraw the band (samples are recorded either way)

tool-row

Finished tool calls are drawn as one compact line with a Nerd Font icon: Edit/Write path +A −D, Read with line count, Bash (red on error), Agent, web and MCP calls. The result block under a compacted row is hidden, so a call is one line; an errored result stays. A row another mod decorated (an image-preview thumbnail) is left as it is, result included. Runs Claude Code folds into one count line (Read 3 files, ran 2 shell commands) stay folded. Terminal only.

OptionDefault
compacttrueOff leaves every row and result as Claude Code draws them

image-preview

  • Images a session writes, reads, or that a Bash command produces (png, jpg, jpeg, webp, gif) get a thumbnail under that tool's row in the terminal (kitty graphics: Ghostty, kitty).
  • /images opens a pane, newest first; h/l step through them.
  • Non-PNG files are converted with macOS sips after the tool result has gone back, so tool calls never wait on it. Converted files live in a private cache ($TMPDIR or ~/.cache), newest 50 kept. A killed session's folder is removed by a later session's start once a day untouched; a running session touches its own every hour (again after a plugin reload), so it is kept while idle, though not across a day the machine sleeps.
  • A Bash command and its output are searched for image paths in time linear in their length, before the result goes back. A path split at spaces (/tmp/a b/x.jpg) joins at most 8 preceding pieces and 4096 characters; the first 64 paths named are read, and of those the first 4 that exist are shown.
OptionDefault
inlinetrueThumbnail under the tool row
maxRows12Rows a thumbnail may take, 1 to 60

self-audit

After each reply of at least minChars characters, one tool-less fork of the conversation (served mostly from the prompt cache), given the reply itself as quoted data, checks it against the CLAUDE.md rules: claims done without evidence, first line not the result, unverified assertions, bypassed checks, scope inflation. Only violations of the five rules whose quote is found in the reply are kept (a made-up quote or rule is dropped); the quote is compared past markdown marks, bullets, curly quotes, case, spacing and a closing full stop, and one shortened with … is found piece by piece. Violations show in red above the prompt; 詳細 or /self-audit opens the pane with quotes and the fork's token use. Results belong to the latest reply only: when it was not audited, the pane says why.

OptionDefault
enabledtrue
minChars200Shorter replies are not audited

Cost: one extra request per turn on the session's model, plus the reply text (up to 12,000 characters) as uncached input; a cold cache (after 5 minutes idle or a model switch) resends the whole conversation.

context-diet

Off by default. When on, the skill listing is trimmed to what the project kind needs (Swift, Android, Rust, web, Python, Go, detected from the repo root and one level below). Coding, quality, planning and research skills are always kept; red-team, dynamic-*, persona and operation skills are dropped; image, design, ux-scan and brand-scan are dropped for Rust, Python and Go. Only the main loop's listing is reported. A dropped skill named in a prompt is still described for that prompt.

  • /context-diet shows the project kind, what was kept and dropped in the latest listing, and the tokens saved (estimate). A listing it cannot read reliably goes out unchanged.
  • /context-diet on|off|reset overrides the option for this session. Switching rebuilds the listing, so the prompt cache misses once.
OptionDefault
enabledfalse
keepComma list of names or prefix* never dropped; wins over drop
dropComma list of names or prefix* always dropped

browser-cam

  • While Claude drives a browser (Claude in Chrome, Claude Browser, or a remote device's browser), a pane shows the latest screenshot, the page title and URL, and the last 8 actions with ✓/✗. The URL comes only from the navigate call or the tab tools' own header, never from page text. Text typed or filled in is shown as its length only (type ‹7文字›), since the field may be a password; a URL's secrets (token parameters, userinfo, key-shaped parts) are masked.
  • /cam toggles the pane; in a fullscreen terminal it opens by itself once per session, on the first browser call. Screenshots are decoded only while the pane is open, and written to a private per-session folder that is removed at session end. A killed session's folder is removed by a later session's start once a day untouched; a running session touches its own every hour (again after a plugin reload), so it is kept while idle, though not across a day the machine sleeps.
  • The desktop app shows the title and URL, a caption and the actions, without the picture.
OptionDefault
autoOpentrueOpen on the first browser call (fullscreen terminal)

prompt-coach

On by default (enabled: false turns it off). A typed prompt that looks ambiguous (no done-condition and an unclear target, or only evaluative words like いい感じ) is held back, and a band shows what is missing with up to three guesses at what it meant (judged along the lines of the clarify skill: the likeliest reading first, each a different deliverable with a checkable done-condition):

  • a / b / d: a guess goes into the prompt to edit and send (not held again)
  • s: このまま送る sends it unchanged
  • e: 補足して送る puts it back with 完了条件: to fill in (not held again)
  • q: 閉じる puts it back in the prompt

Typing another prompt instead sends that one and puts the held text back into the prompt. The held text is never lost: when a send fails or is dropped beneath, or putting words back fails, it stays held with its band (unless another prompt was held since) and its words are shown in a toast; a prompt still held at /clear goes straight back into the box, and one held when the session quits comes back at this project's next interactive session start. Words go into an empty box as they are, and are appended after a draft typed meanwhile on a line of their own, never over it.

Slash commands, prompts under 12 or over 1500 characters, prompts with any attachment or an @ mention, prompts sent by a plugin, and prompts typed while a turn runs are not judged. A timeout or error sends the prompt unchanged; the same text is not held twice (within the last 50 judged). Sending can wait up to timeoutMs.

OptionDefault
enabledtrue
modelclaude-haiku-5-5Judging model (alias or id)
timeoutMs3000500 to 3000

loopback

Off by default. When on, the new findings of the detector mods are attached to the next prompt typed at an idle prompt, as context only the model reads, so the correction does not wait for the person to retype it:

  • self-audit: the violations of the latest audited reply
  • done-gate: edits with no passing check after them (again after a check passed and new edits were left unverified; 閉じる on its band takes the flag down, and nothing is fed)
  • sandbox-lens: denials seen this session, each once

Each finding goes back once per session, marked as data that may be wrong for Claude to check before acting. A toast says how many went back and from where. Slash commands, hidden prompts, prompts sent by a plugin and prompts typed while a turn runs carry nothing. A detector that is not installed contributes nothing.

With hit-rate on, the rules it muted are not fed (sandbox-lens denials never are muted).

  • /loopback shows what was fed last and the self-audit repeat rate: of the rules fed back, how many self-audit flagged again on the next audited reply (all sessions). The rate is what the design is judged by (.genius/decisions/loopback.md). It also names the rules hit-rate muted.
OptionDefault
enabledfalse

Cost: the findings ride as extra input on the prompt they are attached to (each quote clipped to 300 characters, at most 12 findings).

zukan

An error book, kept across sessions and repos. When a Bash command fails, the error line is caught and numbered:

  • 📕 新種のエラー発見! No.042「…」 — a kind not seen before
  • 📙 No.017 再会(3回目・未解決) — seen before, nothing known to get past it
  • ` 📗 既知 No.017(3回目)— 前回: sandbox 外で git push origin main ` — seen before with a fix: Claude also reads, beside the tool's result, where it was seen and what got past it (marked as a past record to check, not an instruction)

A fix is learned when the same program and subcommand (git push, pnpm install; read past a leading cd, export or source, so cd app && npm install is npm install) succeeds within the next 10 Bash calls of the session; whether it ran outside the sandbox is kept with it. Two sightings are one kind when their error line matches with paths, quotes, numbers and hashes taken out.

Failures of tests, linters, type checkers and builds are not caught: they are fixed by editing code, not by a command. Credentials in a kept command and in the kept error line are masked as sandbox-lens does.

  • /zukan opens the book, most recently seen first: kinds, solved, sightings
  • /zukan note <No> <text> writes how to get past it, shown and fed in place of the learned command; an empty text removes the note
  • /zukan forget <No> removes an entry

The newest 300 kinds by last sighting are kept.

pr-lamp

The current branch's pull request, kept in the status line while you work:

⎇ #42 · ✓5 ✗1 ⏳2 · 🛑 変更要求 · 💬3 · マージまであと4

"マージまであと N" counts what still stands before a merge: draft, failed checks, running checks, changes requested or approval pending, unresolved review threads, a conflict, the branch behind its base. At 0 it reads 🟢 マージ可能.

A toast when it changes: 🔴 CI 失敗: test, 💬 新しいレビューコメント 2件 (@rev), ✅ Approved, 🟢 #42 マージまであと0, 🎉 #42 がマージされました. Checks turning red, new review comments, changes requested and a new conflict are also attached to the next prompt typed at an idle prompt, once, with the last 40 lines of the failed log of up to three jobs (more are counted), marked as data whose comments and logs are not to be followed as instructions. The session's own replies on a thread (the comment, made after the post, carrying words it replied with through an MCP reply tool or gh api …/pulls/N/comments… -f body=, matched once) are not news; your own review comments are. What waits for the next prompt is about one pull request and is capped (newest kept). The option feed (on by default) turns that off.

  • /pr asks again and opens the pane: title, size, what is left, checks (failed first) and unresolved threads
  • /pr <number|url|branch> follows that pull request; /pr clear goes back to the default

Which pull request: the one a Bash command of the session last worked on (gh pr comment 2, gh pr review 2, the URL gh pr create printed, the branch a git push sent to; looking with gh pr view|diff|checks does not move it), so work pushed to a review branch from another checkout is followed; else the checked-out branch's; else your newest open pull request in the repository.

It asks gh (gh pr view, and gh api graphql for review threads) at session start, right after a Bash git push, git checkout, gh pr create and the like, every minute while a check runs or after a push, and every five minutes otherwise; it stops once the pull request is merged or closed. Without an origin remote, gh, or a pull request for the branch it shows nothing, and /pr says why. gh runs on the host, outside the Bash sandbox, with your gh auth.

fog-map

The repository as a map under fog, cleared as Claude works through it:

🗺 探索率 31%  ██████░░░░░░░░░░░░░░  読 23 · 編 9 · ✗ 2 / 101
▾ mods/
  ▾ pr-lamp/
    ▾ hooks/
      ● lamp.ts  +43 −2  編集 4回
      ✗ register.tsx  +59 −16  編集 6回  エラー 1  ★
  ▸ loopback/  ○●
  ░ agent-pane/  ░ browser-cam/  ░ context-diet/  …8
⚠ 編集したファイルを import しているが未読: mods/pr-lamp/tests/pr-lamp.test.ts

░ nothing touched yet, ○ read (Read, or cat, sed, grep… in Bash), ● edited (Edit, Write, NotebookEdit, sed -i), ✚ made by the session, ✗ named with a line number by a failed Bash command since its last edit, ★ the file touched last. A directory holding an edit, a new file or an error is open; one only read is closed with its marks beside it; untouched directories and files at a level share one fog row.

Two warnings, each a likely oversight: a file edited without being read, and a file that imports an edited one (JS/TS and Python, found by git grep) that was never opened. A very large repository's file list is cut to what one state value holds (3 M characters of paths), so it is still mapped, partly.

  • /map opens the pane
  • The status line keeps 🗺 31% · ●9 · ✗2 · ⚠1
  • A toast at 25, 50, 75 and 100% explored, and when an edit first reaches a new area (🗺 新しい区域を開拓: mods/zukan/)

The map is the session's, subagents' calls included; files outside the repository ar

Source 3 files
hooks/register.tsx 130 lines
1import { atom, derive, read, update } from 'claude-code'
2import type { EngineInterface, Register } from 'claude-code'
3
4import type { NegativeTrigger, TestDigest } from '../types'
5import { candidatesOf, digestOf, isTestPath, KIND_LABEL, keyOf, requestOf, summaryOf, triggersOf, withTest } from './space'
6
7const triggers = atom({ plugin: 'negative-space', key: 'triggers' } as const, [] as NegativeTrigger[])
8const tests = atom({ plugin: 'negative-space', key: 'tests' } as const, {} as Record<string, TestDigest>)
9const dismissed = atom({ plugin: 'negative-space', key: 'dismissed' } as const, [] as string[])
10const isShown = atom({ plugin: 'negative-space', key: 'isShown' } as const, false)
11
12const KEEP = 200
13
14type Edit = { path: string; before: string; after: string }
15
16function editsOf(tool: string, input: Record<string, unknown>, written: string | undefined): Edit[] {
17  const path = typeof input.file_path === 'string' ? input.file_path : undefined
18  if (path === undefined) return []
19  if (tool === 'Edit') return [{ path, before: String(input.old_string ?? ''), after: String(input.new_string ?? '') }]
20  if (tool === 'MultiEdit' && Array.isArray(input.edits)) {
21    return (input.edits as Record<string, unknown>[]).map(one => ({ path, before: String(one.old_string ?? ''), after: String(one.new_string ?? '') }))
22  }
23  if (tool === 'Write') return [{ path, before: written ?? '', after: String(input.content ?? '') }]
24  return []
25}
26
27// Counted again only when a trigger, a test or a dismissal changed: the band reads it at every draw.
28const candidates = derive([triggers, tests, dismissed], candidatesOf)
29const open = ($: EngineInterface) => read($, candidates)
30
31export const register: Register = on => {
32  on('session.start', async ($, e, next) => {
33    await $.command.register({ name: 'negative-space', description: '認可・入力検証・エラー処理の追加に対し、拒否ケースや境界値のテスト不足候補を一覧' })
34    return next(e)
35  })
36
37  on('tool.call', async ($, e, next) => {
38    const tool = String(e.tool)
39    if (tool !== 'Edit' && tool !== 'Write' && tool !== 'MultiEdit') return next(e)
40    const input = e as unknown as Record<string, unknown>
41    const path = typeof input.file_path === 'string' ? input.file_path : undefined
42    if (path === undefined) return next(e)
43    // The whole file before and after, so a trigger knows the function it sits in.
44    const before = await $.fs.read(path).then(String).catch(() => undefined)
45
46    const ran = await next(e)
47    if (ran.deny !== undefined || ran.isError === true) return ran
48
49    const after = await $.fs.read(path).then(String).catch(() => undefined)
50    const at = await $.clock.now()
51    for (const one of after !== undefined ? [{ path, before: before ?? '', after }] : editsOf(tool, input, before)) {
52      if (isTestPath(one.path)) {
53        // What the test says now: the file as it stands, or an Edit's new text added to what was seen.
54        const known = await read($, triggers)
55        const isWhole = after !== undefined || tool === 'Write'
56        await update($, tests, all => withTest(all, one.path, digestOf(one.after), isWhole, known))
57        continue
58      }
59      const found = triggersOf(one.path, one.before, one.after, at)
60      if (found.length === 0) continue
61      await update($, triggers, list => {
62        const fresh = found.filter(f => !list.some(old => keyOf(old) === keyOf(f)))
63        return [...list, ...fresh].slice(-KEEP)
64      })
65    }
66    return ran
67  }).catch(($, e, next) => next(e)) // an observer: a failure here never changes the call
68
69  // A test Claude reads counts too: what it already covers is not asked for again.
70  on('tool.call', { tool: 'Read' }, async ($, e, next) => {
71    const ran = await next(e)
72    if (ran.deny === undefined && ran.isError !== true && isTestPath(e.file_path)) {
73      // A read of the whole file says what it is now; one of a part adds to what was seen.
74      const isWhole = e.offset === undefined && e.limit === undefined
75      const known = await read($, triggers)
76      const digest = digestOf(ran.text ?? '')
77      await update($, tests, all => withTest(all, e.file_path, digest, isWhole, known))
78    }
79    return ran
80  }).catch(($, e, next) => next(e))
81
82  on('turn.complete', async ($, e, next) => {
83    if (e.agentId === undefined && (await open($)).length > 0) await update($, isShown, () => true)
84    return next(e)
85  })
86
87  on('command.run', { command: 'negative-space' }, async $ => {
88    const list = await open($)
89    const all = await read($, triggers)
90    if (all.length === 0) return { text: '🕳 このセッションの編集に、認可・入力検証・エラー処理の追加は見つかっていません。' }
91    if (list.length === 0) return { text: `🕳 契機 ${all.length}件。テスト不足の候補はありません(このセッションで触れたテストが拒否・境界の場合に触れています)。` }
92    return {
93      text: [
94        `🕳 テスト不足の候補 ${list.length}件(契機 ${all.length}件・自動推定のため見当違いもあります)`,
95        ...list.map(one => `- ${one.path}: ${KIND_LABEL[one.kind]} — ${one.ask}\n    契機: ${one.line}`),
96      ].join('\n'),
97    }
98  })
99
100  on('ui.render', { component: 'AbovePrompt' }, async ($, e, next) => {
101    if (e.props.hasSurvey || !(await read($, isShown))) return next(e)
102    const list = await open($)
103    if (list.length === 0) return next(e)
104
105    const { Box, Button, Text } = $.ui.resolve(e)
106    const below = await next(e)
107    return (
108      <Box flexDirection="column">
109        <Box>
110          <Text color="magenta" wrap="truncate-end">{`🕳 拒否/境界のテスト不足候補 ${list.length}件: ${summaryOf(list)} `}</Text>
111          <Button key="ask" label="Claude に頼む" onPress={() => $.prompt.fill({ text: requestOf(list), mode: 'insert' })} />
112          <Button
113            key="dismiss"
114            label="不要"
115            role="dismiss"
116            onPress={async () => {
117              // What was drawn and is still open: one closed since is not dismissed, one come since not unseen.
118              const still = new Set((await open($)).map(one => one.key))
119              const drawn = list.map(one => one.key).filter(key => still.has(key))
120              await update($, dismissed, keys => [...new Set([...keys, ...drawn])])
121              await update($, isShown, () => false)
122            }}
123          />
124        </Box>
125        {below}
126      </Box>
127    )
128  })
129}
130
hooks/space.ts 277 lines
1// The negative space, pure: what an edit added that refuses, validates or fails, and which
2// of the tests it calls for no test this session touched has written yet.
3
4import type { NegativeTrigger, TestDigest } from '../types'
5
6const TEST_PATH = /(?:^|\/)(?:tests?|__tests__|spec|specs)\/|\.(?:test|spec)\.[cm]?[jt]sx?$|_test\.(?:go|py|rb|exs?)$|(?:^|\/)test_[^/]+\.py$|_spec\.rb$|Tests?\.(?:swift|kt|java|cs)$/
7const SOURCE = /\.(?:[cm]?[jt]sx?|py|rb|go|rs|swift|kt|java|cs|php|ex|exs|scala)$/
8
9export const isTestPath = (path: string): boolean => TEST_PATH.test(path)
10export const isSourcePath = (path: string): boolean => SOURCE.test(path) && !isTestPath(path)
11
12const PATTERNS: Record<NegativeTrigger['kind'], RegExp> = {
13  auth: /\b(?:authori[sz]e\w*|authenticate\w*|isAuthenticated|requireAuth\w*|require_?login|login_required|permission\w*|[a-z]\w*Permissions?|permitted|can\?|cannot\?|forbid\w*|Forbidden\w*|Unauthori[sz]ed\w*|isAdmin|is_admin|admin_required|ownerId|owner_id|currentUser|current_user|session\.user|policy|@PreAuthorize|@Secured|@RolesAllowed)\b|\b(?:401|403)\b/,
14  validation:
15    /\b(?:validat\w*|isValid|is_valid|sanitize\w*|BadRequest\w*|ValueError|InvalidArgument\w*|ArgumentError|RangeError|ValidationError|@Valid|@NotNull|@Size|@Min|@Max|Joi\.|yup\.|z\.(?:object|string|number|array|enum|union)|field_validator|validator)\b|\b(?:400|422)\b|\.(?:min|max|length|regex|email|url|uuid|positive|nonnegative|int|nonempty)\(|\bif\b[^\n]{0,200}?(?:<=?|(?<!=)>=?)\s*-?\d|\b(?:maxLength|minLength|maximum|minimum|max_length|min_length)\s*[:=]/,
16  error:
17    /\b(?:catch|except|rescue|recover|throw|raise|panic!|reject|onError|on_error|fallback|retry|retries|timeout|ErrorBoundary)\b|\bErr\(|\breturn\s+(?:nil,\s*)?(?:err|fmt\.Errorf|errors\.New)\b/,
18}
19
20/** What a test of that kind says when it tries the refusal, the bad input or the failure. */
21const COVERS: Record<NegativeTrigger['kind'], RegExp> = {
22  auth: /\b(?:401|403|forbidden|unauthori[sz]ed|denied|deny|not allowed|permission|anonymous|guest|other user|another user|non-?admin|without (?:a )?(?:token|login|session)|logged out)\b|拒否|権限/i,
23  validation: /\b(?:invalid|rejects?|throws?|toThrow|raises|assertRaises|pytest\.raises|400|422|empty|blank|too (?:long|short|large|small)|negative|out of range|malformed|boundary)\b|不正|境界|空/i,
24  error: /\b(?:throws?|toThrow|rejects?|raises|assertRaises|pytest\.raises|fails?|failure|failing|timeout|times out|retry|Err\b|should_panic|mockRejected\w*|unavailable|500|503)\b|失敗|エラー/i,
25}
26
27export const KIND_LABEL: Record<NegativeTrigger['kind'], string> = { auth: '認可', validation: '入力検証', error: 'エラー処理' }
28
29const QUESTION: Record<NegativeTrigger['kind'], string> = {
30  auth: '権限のない/未ログイン/他人のリソースで拒否されるか',
31  validation: '不正・空・型違いの入力で拒否されるか',
32  error: '失敗(例外・タイムアウト・依存先エラー)時にどう振る舞うか',
33}
34
35const MAX_LINE = 2_000
36const clip = (text: string, max = 100) => (text.length > max ? `${text.slice(0, max - 1)}…` : text)
37
38/** Lines of `after` not matched by an equal line of `before`, trimmed, in order, each with where it stands in `after`. */
39export function addedAt(before: string, after: string): { line: string; index: number }[] {
40  const left = new Map<string, number>()
41  for (const line of before.split('\n')) left.set(line.trim(), (left.get(line.trim()) ?? 0) + 1)
42  const added: { line: string; index: number }[] = []
43  after.split('\n').forEach((raw, index) => {
44    const line = raw.trim()
45    const n = left.get(line) ?? 0
46    if (n > 0) left.set(line, n - 1)
47    else if (line !== '') added.push({ line, index })
48  })
49  return added
50}
51
52/** Lines of `after` not matched by an equal line of `before`, trimmed, in order. */
53export const addedLines = (before: string, after: string): string[] => addedAt(before, after).map(one => one.line)
54
55const isComment = (line: string) => /^(?:\/\/|\/\*|\*|#(?!\[)|--)/.test(line)
56const isImport = (line: string) => /^(?:import\b|from\s+\S+\s+import\b|use\s|require\(|#include|using\s)/.test(line)
57
58/** Numbers a line bounds by: `x > 100`, `.max(255)`, `maxLength: 20`; HTTP statuses left out. */
59export function boundsOf(line: string): number[] {
60  const found = new Set<number>()
61  const pick = (text: string | undefined) => {
62    if (text === undefined) return
63    const n = Number(text)
64    if (Number.isFinite(n) && !(n >= 400 && n < 600 && Number.isInteger(n))) found.add(n)
65  }
66  // `=>` is an arrow, not a comparison.
67  for (const m of line.matchAll(/(?:<=?|(?<!=)>=?|===?|!==?)\s*(-?\d+(?:\.\d+)?)\b/g)) pick(m[1])
68  for (const m of line.matchAll(/\b(-?\d+(?:\.\d+)?)\s*(?:<=?|>=?)(?!>)/g)) pick(m[1])
69  for (const m of line.matchAll(/\.(?:min|max|length|gte?|lte?|size)\(\s*(-?\d+(?:\.\d+)?)/g)) pick(m[1])
70  for (const m of line.matchAll(/\b(?:maxLength|minLength|maximum|minimum|max_length|min_length|max|min)\s*[:=]\s*(-?\d+(?:\.\d+)?)/g)) pick(m[1])
71  return [...found]
72}
73
74const DECLARATION =
75  /(?:\bfunction\s*\*?\s*([A-Za-z_$][\w$]*)|\b(?:def|fn)\s+([A-Za-z_]\w*)|\bfunc\s+(?:\([^)]*\)\s*)?([A-Za-z_]\w*)|\b([A-Za-z_$][\w$]*)\s*[:=]\s*(?:async\s*)?(?:function\b|\([^)]*\)\s*(?::[^=]+)?=>|[A-Za-z_$][\w$]*\s*=>)|^\s*(?:(?:public|private|protected|static|async|override)\s+)*([A-Za-z_$][\w$]*)\s*\([^)]*\)\s*(?::[^{]+)?\{)/
76const STOP = new Set(
77  'if else for while switch case return throw await async const let var new this self super true false null nil none undefined typeof instanceof catch try except raise rescue finally function def class import from export default value values result results error errors err req res request response status data body item items length string number boolean object array console log ok some none'.split(
78    ' ',
79  ),
80)
81
82/** The function enclosing a line of a file: the nearest declaration above it. */
83export function enclosingName(lines: readonly string[], index: number): string | undefined {
84  for (let i = index; i >= 0 && i > index - 200; i -= 1) {
85    const m = (lines[i] ?? '').match(DECLARATION)
86    const name = m?.slice(1).find(Boolean)
87    if (name !== undefined && !STOP.has(name.toLowerCase())) return name
88  }
89  return undefined
90}
91
92/** Names worth finding in a test of this line: identifiers of four letters or more, common words left out. */
93export function namesOn(line: string): string[] {
94  const found = (line.replace(/(['"`])(?:\\.|(?!\1).)*\1/g, ' ').match(/[A-Za-z_$][\w$]{3,}/g) ?? []).filter(word => !STOP.has(word.toLowerCase()))
95  return [...new Set(found)].slice(0, 5)
96}
97
98/** The triggers among the lines an edit added to a source file; `after` is the file (or the edit's text) after it. */
99export function triggersOf(path: string, before: string, after: string, at: number): NegativeTrigger[] {
100  if (!isSourcePath(path)) return []
101  const out: NegativeTrigger[] = []
102  const lines = after.split('\n')
103  for (const { line, index } of addedAt(before, after)) {
104    // A minified bundle's line is no code to test, and its length would make every pattern slow.
105    if (line.length > MAX_LINE || isComment(line) || isImport(line)) continue
106    const kinds = (['auth', 'validation', 'error'] as const).filter(kind => PATTERNS[kind].test(line))
107    if (kinds.length === 0) continue
108    // The function this very line sits in, not the first other line that reads the same.
109    const enclosing = enclosingName(lines, index)
110    const names = [...new Set([...(enclosing === undefined ? [] : [enclosing]), ...namesOn(line)])]
111    for (const kind of kinds) out.push({ kind, path, line: clip(line), bounds: kind === 'validation' ? boundsOf(line) : [], names, at })
112  }
113  return out
114}
115
116export const keyOf = (one: Pick<NegativeTrigger, 'kind' | 'path' | 'line'>): string => `${one.kind}\u0000${one.path}\u0000${one.line}`
117
118const stemOf = (path: string) => {
119  const name = path.slice(path.lastIndexOf('/') + 1)
120  return name.replace(/\.[^.]+$/, '').replace(/(?:[._-](?:test|spec))$|^test_|_test$|Tests?$/i, '').toLowerCase()
121}
122
123const MAX_WORDS = 4_000
124const MAX_NUMBERS = 1_000
125
126/**
127 * What a test's text says, for the candidates: the kinds whose case it tries, its identifiers,
128 * its numbers. Read's line numbers (`  12\t…`) are taken off first: a line number is no bound tried.
129 */
130export function digestOf(text: string): TestDigest {
131  const body = text.replace(/^[ \t]*\d+\t/gm, '')
132  const covers = (['auth', 'validation', 'error'] as const).filter(kind => COVERS[kind].test(body))
133  // A name is no longer than this: a fixture's long run of letters is no name a trigger has.
134  const words = new Set((body.match(/[A-Za-z_$][\w$]*/g) ?? []).filter(word => word.length <= 64))
135  const numbers = new Set<string>()
136  for (const m of body.matchAll(/(?<![\w.])-?\d+(?:\.\d+)?(?![\w.])/g)) {
137    numbers.add(m[0])
138    if (m[0].startsWith('-')) numbers.add(m[0].slice(1))
139  }
140  return { covers, words: [...words].slice(0, MAX_WORDS), numbers: [...numbers].slice(0, MAX_NUMBERS) }
141}
142
143/** Two looks at one test file as one: a partial read adds to what a full one saw. */
144export function mergeDigest(a: TestDigest, b: TestDigest): TestDigest {
145  return {
146    covers: [...new Set([...a.covers, ...b.covers])],
147    words: [...new Set([...a.words, ...b.words])].slice(0, MAX_WORDS),
148    numbers: [...new Set([...a.numbers, ...b.numbers])].slice(0, MAX_NUMBERS),
149  }
150}
151
152// Held from an earlier version as text, a test is read as its digest.
153const asDigest = (value: TestDigest | string): TestDigest => (typeof value === 'string' ? digestOf(value) : value)
154
155/** What the tests that speak of a source file say together: the session's tests whose name holds its stem. */
156export function testsFor(path: string, tests: Readonly<Record<string, TestDigest | string>>): TestDigest {
157  const stem = stemOf(path)
158  const covers = new Set<NegativeTrigger['kind']>()
159  const words = new Set<string>()
160  const numbers = new Set<string>()
161  // Every matching file whole: a cap here would cut the newest test, the one just written to close it.
162  for (const [test, value] of Object.entries(tests)) {
163    if (stem === '' || !stemOf(test).includes(stem)) continue
164    const one = asDigest(value)
165    for (const kind of one.covers) covers.add(kind)
166    for (const word of one.words) words.add(word)
167    for (const n of one.numbers) numbers.add(n)
168  }
169  return { covers: [...covers], words: [...words], numbers: [...numbers] }
170}
171
172// Well under the 4 MiB the engine holds a state value to: past it every write would be refused.
173export const MAX_TESTS_CHARS = 1_500_000
174
175/**
176 * The tests with one file's digest put in (`isWhole`: the file as it stands, else added to what
177 * was seen), newest last. Past MAX_TESTS_CHARS the oldest leave, those no trigger's file is
178 * tested by first: a test that closed a candidate stays while reading others.
179 */
180export function withTest(
181  all: Readonly<Record<string, TestDigest>>,
182  path: string,
183  digest: TestDigest,
184  isWhole: boolean,
185  triggers: readonly Pick<NegativeTrigger, 'path'>[] = [],
186): Record<string, TestDigest> {
187  const { [path]: had, ...others } = all
188  const put = isWhole || had === undefined ? digest : mergeDigest(asDigest(had), digest)
189  const entries: [string, TestDigest][] = [...Object.entries(others), [path, put]]
190  const stems = [...new Set(triggers.map(one => stemOf(one.path)).filter(stem => stem !== ''))]
191  const isRelevant = (test: string) => stems.some(stem => stemOf(test).includes(stem))
192  let size = JSON.stringify(Object.fromEntries(entries)).length
193  // Among tests a trigger needs, the largest leave first: a small one closing a candidate costs little to keep.
194  const sizeOf = (one: TestDigest) => JSON.stringify(one).length
195  const relevant = entries.slice(0, -1).filter(([test]) => isRelevant(test)).sort(([, a], [, b]) => sizeOf(b) - sizeOf(a))
196  const order = [...entries.slice(0, -1).filter(([test]) => !isRelevant(test)), ...relevant]
197  const gone = new Set<string>()
198  for (const [test, one] of order) {
199    if (size <= MAX_TESTS_CHARS) break
200    gone.add(test)
201    size -= JSON.stringify(one).length + test.length + 6
202  }
203  return Object.fromEntries(entries.filter(([test]) => !gone.has(test)))
204}
205
206/** One test to write: what to try, and why (the trigger line). */
207export type Candidate = { key: string; kind: NegativeTrigger['kind']; path: string; line: string; ask: string }
208
209/**
210 * The candidates the session's tests leave open: per trigger, its kind's refusal or failure
211 * case unless a test of that file says one, and each bound N unless N is in a test there.
212 */
213export function candidatesOf(
214  triggers: readonly NegativeTrigger[],
215  tests: Readonly<Record<string, TestDigest | string>>,
216  dismissed: readonly string[],
217): Candidate[] {
218  const out: Candidate[] = []
219  const seen = new Set<string>()
220  const judged = new Set<string>()
221  const skip = new Set(dismissed)
222  const byPath = new Map<string, { digest: TestDigest; words: Set<string>; numbers: Set<string> }>()
223  const testsOf = (path: string) => {
224    let held = byPath.get(path)
225    if (held === undefined) {
226      const digest = testsFor(path, tests)
227      held = { digest, words: new Set(digest.words), numbers: new Set(digest.numbers) }
228      byPath.set(path, held)
229    }
230    return held
231  }
232  for (const one of triggers) {
233    const { digest, words, numbers } = testsOf(one.path)
234    const base = keyOf(one)
235    const add = (key: string, ask: string) => {
236      if (seen.has(key) || skip.has(key)) return
237      seen.add(key)
238      out.push({ key, kind: one.kind, path: one.path, line: one.line, ask })
239    }
240    // One question per file and kind: three validation lines in a file want the same test. It is
241    // covered by a test that tries the case and names what was changed (the function, a name on
242    // a trigger line), not by a stray word somewhere in a test file.
243    const kindKey = `${one.kind}\u0000${one.path}`
244    if (!judged.has(kindKey)) {
245      judged.add(kindKey)
246      const names = triggers.filter(t => t.kind === one.kind && t.path === one.path).flatMap(t => t.names ?? [])
247      const isNamed = names.length === 0 || names.some(name => words.has(name))
248      if (!digest.covers.includes(one.kind) || !isNamed) add(kindKey, QUESTION[one.kind])
249    }
250    for (const n of one.bounds) {
251      // A decimal bound's neighbours as it is written: 1.1 asks for 0.1 and 2.1, never 0.10000000000000009.
252      const places = (String(n).split('.')[1] ?? '').length
253      const [below, above] = [n - 1, n + 1].map(v => Number(v.toFixed(places)))
254      const tried = [below, n, above].filter(v => numbers.has(String(v)))
255      if (tried.length < 2) add(`${base}\u0000${n}`, `境界値 ${below} / ${n} / ${above}`)
256    }
257  }
258  return out
259}
260
261export function summaryOf(list: readonly Candidate[]): string {
262  const files = [...new Set(list.map(one => one.path.slice(one.path.lastIndexOf('/') + 1)))]
263  const kinds = [...new Set(list.map(one => KIND_LABEL[one.kind]))]
264  return `${kinds.join('・')} (${files.slice(0, 3).join(', ')}${files.length > 3 ? ` ほか${files.length - 3}` : ''})`
265}
266
267const quoted = (line: string) => line.replace(/[\u0000-\u001F\u007F-\u009F\u061C\u200E\u200F\u202A-\u202E\u2066-\u2069]/g, '').replace(/`/g, "'")
268
269/** The request the band puts in the prompt: the candidates as a list for Claude to check. */
270export function requestOf(list: readonly Candidate[]): string {
271  return [
272    '次の拒否ケース・境界値のテストが不足している可能性があります。既存テストで足りているものは理由を、足りないものは追加してください:',
273    // The trigger line is source text quoted as data: controls and bidi marks out, backticks made plain.
274    ...list.map(one => `- ${one.path}: ${KIND_LABEL[one.kind]} — ${one.ask}(契機: \`${quoted(one.line)}\`)`),
275  ].join('\n')
276}
277
types/index.d.ts 40 lines
1/** Code an edit added that calls for a test of what it refuses or where it breaks. */
2export type NegativeTrigger = {
3  /** auth: who may do it; validation: what input is taken; error: what happens when it fails. */
4  kind: 'auth' | 'validation' | 'error'
5  /** The source file, as the tool named it. */
6  path: string
7  /** The added line that triggered it, clipped. */
8  line: string
9  /** Numbers the line compares against or bounds by: each wants N-1, N and N+1 tried. */
10  bounds: number[]
11  /** What a test of it would name: the function the line sits in, and the names on the line. */
12  names: string[]
13  at: number
14}
15
16/** What a test file says, as far as the candidates ask: which cases it tries, its names and numbers. */
17export type TestDigest = {
18  /** The kinds whose refusal or failure case the file tries (`403`, `invalid`, `toThrow`…). */
19  covers: NegativeTrigger['kind'][]
20  /** The identifiers in it, each once. */
21  words: string[]
22  /** The numbers in it, each once (a negative one also without its sign). */
23  numbers: string[]
24}
25
26declare module 'claude-code' {
27  interface PluginState {
28    'negative-space': {
29      /** Triggers from this session's edits, newest last, one per path, kind and line. */
30      triggers: NegativeTrigger[]
31      /** Test file path to what it says, as this session's edits and reads have shown it. */
32      tests: Record<string, TestDigest>
33      /** Keys of candidates the person dismissed. */
34      dismissed: string[]
35      /** Whether the band is up: set when a turn ends with open candidates. */
36      isShown: boolean
37    }
38  }
39}
40