Remove the legacy plan2code-loop CLI

AI Assisted
This commit is contained in:
jparkerweb committed 2026-11-02 13:22:59 -08:00
1 parent d7abcd2ac6
commit c8d12a3d99
89 files changed
+449 -3189

No files matched your search

+10 -6
View File
@@ -29,9 +29,6 @@ plan2code/
│ │ └── ai-agent-file-sync.md # Step 7: replace AI configs with AGENTS.md refs
│ └── plan2code-4-finalize-references/ # Reference files for finalize skill
│ └── community-feedback-submission.md # STEP 6.5 payload schema + submission tiers
├── plan2code-loop/ # Autonomous loop CLI tool (Node.js/TypeScript)
│ ├── src/ # TypeScript source
│ └── dist/ # Built output (tsup)
├── plan2code-bot/ # Autonomous end-to-end workflow test runner (TypeScript, tsup, Claude Agent SDK, vitest); user docs in .readme/test-bot.md
├── plan2code-metrics/ # Recursive self-improvement toolchain
│ ├── src/ # TypeScript source
@@ -71,7 +68,7 @@ plan2code/
├── .claude/ # Repo-local Claude Code config (NOT installed by install.js)
│ └── skills/ # Maintainer-only dev skills, e.g. plan2code-publish/, plan2code-model-update/
├── .github/ISSUE_TEMPLATE/ # bug.yml, idea.yml, config.yml (target of the console's feedback button)
├── .readme/ # User guides linked from README (web-console.md, web-console-design.md, walkthrough.md, autonomous-loop.md, metrics.md, status-line.md, test-bot.md)
├── .readme/ # User guides linked from README (web-console.md, web-console-design.md, walkthrough.md, metrics.md, status-line.md, test-bot.md)
├── docs/ # Landing page `index.html` served at plan2code.jparkerweb.com (GitHub Pages, `CNAME`) and assets
│ └── screenshots/ # Web console screenshots, <shot>-light.webp and <shot>-dark.webp (generated)
├── specs/ # Feature specs (if any in-progress)
@@ -284,6 +281,10 @@ Optional local browser UI shipped with **every** skill. Source lives in `src/web
| `public/activities/ghoul-ride-scenes.js` | The dark ride's scene table as data (12 rooms, lengths, speeds, spawn rows in author coordinates), no imports |
| `public/activities/ghoul-ride-core.js` | The ride's shared basics: the 96 by 54 field's constants, math helpers, the bound state `G` (a live import), `rnd`/`rand` (seeded) and `vrand` (visual) |
| `public/activities/ghoul-ride-draw.js` | The ride's art entry (`rideArt(d, die)`: `field(ctx, v)` draws the whole field by phase, `tick` the cascade's sounds); `ghoul-draw.js` hands over to it when `round.gallery` is set. Parts: `ghoul-ride-gfx.js` (canvas helpers), `ghoul-ride-draw-kinds.js` (animatronics, mini-bosses, the Barker), `ghoul-ride-draw-sets.js` (the 12 sets, far walls, aisle, near layer), `ghoul-ride-draw-hud.js` (car and Planny, HUD, duck strip, reticle, title, end cascade) |
| `public/activities/ghoul-ride-gfx.js` | The ride's drawing helpers: the bound 2D context and the short fill, stroke and shape calls the ride's art is written in |
| `public/activities/ghoul-ride-draw-kinds.js` | The ride's drawn things: every figure, beast, lantern, blade, door, hand, head, mirror, web and boss, a port of the prototype's draw code (`prototypes/ghoul-rush-ride/gallery.html`) |
| `public/activities/ghoul-ride-draw-sets.js` | The ride's rooms drawn: each room's far wall, set pieces front and back, aisle and the near pillars, a port of the prototype's world draw |
| `public/activities/ghoul-ride-draw-hud.js` | The ride's car and the screen around it: Planny in his car, blood and sparks, the lens, the hearts, the progress bar, the room plaque, the aim, the title and the cleared card |
| `public/activities/ghoul-ride-voices.js` | The ride's `LOOKS`, `VOICES`, `EVENT_VOICE`, `LOSE_WORDS`, `LOSE_SOUNDS` and `react` (every `ride<Name>` event plays tone `ghoulRide<Name>`) |
| `public/activities/ghoul-sfx-ride.js` | The ride's synthesized sounds as `tones` (`ghoulRide<Kind>`) and its music `ghoulRideDrone` (`RIDE_SFX`, merged by `ghoul-sfx.js`) |
| `public/activities/ghoul-frozen.js` | Ghoul Rush's Frozen Lake stage pack, its rules (imports only `burrow` from the Graveyard; draws no random number): uses `tick` (thin ice, the blizzard, the Widow's spikes, web and barrels), `place`, `unseen`, `solid`, `kill`, `target`, `land`, `weak`, `cause`, `guard`, `swoops` and `slides`. The generic slippery-ice rule (`iceRun`, room field `ice`) and thin-ice rule (`stepThin`, room field `thin`, `round.thin`), the under-ice hand, snow shade (unseen, footprints), Frozen Ghoul (slides, shatters into shards), Ice Wraith (its dive freezes), Hot Cocoa, the Frost Widow (veil, barrels, weak spot, blizzard, spikes, web, rage) and `FROZEN_ROOMS`. Every constant is exported and documented in its header |
@@ -307,6 +308,9 @@ Optional local browser UI shipped with **every** skill. Source lives in `src/web
| `public/activities/ghoul-powers.js` | Ghoul Rush's power-ups and pickups (no imports: `ghoul.js` hands its rules a kit, `POWER_KIT`, of `landing`, `MAX_LIVES` and the numbers they need): `POWERS` (each power-up's name, line, color, `ms` if timed, drop `weight`, and its rule, the Lantern's `dark`), `TIMED`, `PILL_WORDS`, `FLAP_CLEAR`, `HOURGLASS_MS`, `CANDY_POINTS`, `POWER_POINTS`, `CRYSTAL_POINTS`, `powerHere(round, kind)` (whether a power-up can turn up in the room as it is now; a new rule, such as a stage-only power-up, goes in `POWERS` and is checked here), `pickPower`, `applyPower(round, kind, kit)` (`ghoul.js` exports it bound to its kit as `applyPower(round, kind)`) and `pickups` (candy, placed and dropped power-ups, crystals up high and the key, taken by touch; drops fall and fade). Split from `ghoul.js` to keep it under the 3,000-line cap |
| `public/activities/ghoul-arms.js` | Ghoul Rush's secondary weapons (imports only `ghoul-powers.js`; a pack in `PACKS`, `ARMS`): the double tap (`pressFire`, `releaseFire`, called from the game's `key()` and `release()`, never from `round.input.fire`, so bots never throw), the bag `round.arms` (`useArm`, `armOnTop`: a ride room skips what needs a floor), each weapon's rule (a shot with `arm` through `moveShots`, or a hazard), and `armBossDamage` (scaled to depth, capped at `ARM_BOSS_SHARE` of a boss's hp a fight). Weapons' names, bundles and caps are in `POWERS` (`arm: true`, weight 0), the stage's own in `STAGE_ARMS`, the kill drop band in `armDrop`, all in `ghoul-powers.js`. |
| `public/activities/ghoul-arms-draw.js` | The weapons' art (`armsArt(d)`): their badges (`powers`), `hazards`, `shot` for one in flight, `pill` for the HUD's icon and count, `front` (stars over the dazed), `lights` and `react`. Dynamite's stick, blast and badge stay in `ghoul-mine-draw.js`. |
| `public/activities/ghoul-rockets.js` | The Bottle Rocket's rules, the common weapon every stage sells: one use lights three rockets, one ahead, one up and one behind |
| `public/activities/ghoul-rockets-draw.js` | The Bottle Rocket's drawing (`rocketArt(d)`): the muzzle's flash and smoke, the flared rocket on a stick with a glowing head, the fountain of sparks it sheds and the burst when it goes off |
| `public/activities/ghoul-pans.js` | The Cast-Iron Skillet's rules, the East Wing's weapon: three pans that circle Planny, then spiral outward until they are off screen |
| `public/activities/ghoul-sfx-arms.js` | The weapons' `tones` (`ARMS_SFX`), merged by `ghoul-sfx.js` like a stage's. A first pass, not yet picked by ear. |
| `scripts/test-web-console-activities-ghoul-arms.mjs` | The weapons' suite: the double tap through the keys (repeats, feathering, hammering), the bag and its pill, each weapon, bosses (scale, cap, guards, the half-hp drop), placement in every room, the drop band's single draw, ride rooms, drawing and sounds. |
| `public/activities/ghoul-run.js` | Ghoul Rush's run flow (no imports: `ghoul.js` hands its rules `RUN_KIT`, the room and stage tables, the timings and the helpers that set a room up, and binds them): `newRound`, `setupRoom`/`resetRoom`, the stage pick (`stagesLeft`, `nextRoom`, `steerPick`, `pickStage`), the stage select for testing (`openSelect`, `steerSelect`, `takeSelect`: a practice round, `round.practice`, that keeps no best score or record), `camera`, `respawn`, `lockArena`, `clearRoom`, and a room's monster supply (`spawners`, `horde`, `hurry`). `ghoul.js` re-exports `newRound`, `stagesLeft`, `steerPick`, `pickStage`, `steerSelect` and `takeSelect`, and its game answers `stageSelect()`. Split from `ghoul.js` to keep it under the 3,000-line cap, with the same random draws in the same order |
@@ -461,7 +465,7 @@ Key design points:
| `state.json` → `helpers` | `console.mjs post` | the helpers the agent reports, merged by `id`, checked by `checkHelpers()` plus the prose scan; dropped by `switchWorkflow()` on a dashboard hop |
`POST /subagents` (origin-checked; `400` `not-offered` / `bad` / `too-long` / `max`) saves the store and bumps the session setting. `consume()`, `cmdChat()`, `deliverChat()` and `open` append `blockFor()`'s lines last in `reply` and the `subagents` field; a change alone never ends a `wait` slice. The page's Subagents tab (before Ask) draws `state.helpers` through `helperRows()`. Helpers never reach the meter ledger. Contract: `console-subagents.md`.
- **The filesystem is the transport.** Sessions live in `~/.plan2code/console/sessions/<sid>/`, deliberately not under `specs/`, so the console cannot trip the metrics scraper or the loop's marker parser, and survives `git clean`. A result written after the agent dies is picked up by the next session.
- **The filesystem is the transport.** Sessions live in `~/.plan2code/console/sessions/<sid>/`, deliberately not under `specs/`, so the console cannot trip the metrics scraper or any marker parser, and survives `git clean`. A result written after the agent dies is picked up by the next session.
- **Attachments on notes and quick questions live in `<sid>/uploads/`**, never in the project: images as `<uuid>.jpg`, documents as `<uuid>.<ext>` with `ext` from `DOC_TYPES` in `answers.js` (text, code, PDF). The browser re-encodes images to JPEG and sends raw bytes to `POST /upload` as `image/jpeg`; documents go as-is as `application/octet-stream`, the extension taken by `docExt()` from the uncut `x-p2c-name`. `readRaw(req, limit, cb)` in `routes.mjs` caps images at `MAX_BODY` (4 MB) and documents at `MAX_DOC_BYTES` (10 MB), and drains rather than destroys so the client gets its 413. Images need the `FF D8 FF` magic; documents pass `checkDocument()` in `sessions.mjs` (`%PDF-` for PDF, strict UTF-8 with no NUL for text), else 415. The server names the file with `crypto.randomUUID()`. `GET /uploads/<id>.<ext>` serves images as `image/jpeg` and documents under a fixed type (`application/pdf`, else `text/plain`) with `nosniff` and, on text, CSP `sandbox` (a sandboxed PDF cannot open in the viewer), so a browser can show but never run one; `recordSubmitted()` writes every send's note onto its item as `sentNote { at, text, images, files }` (paths and names only), which `noteAwaiting()` in `answers.js` turns into the Sent box's attachment links and the Notes pane's pending note until the agent's next thread message; `DELETE /uploads/<id>.<ext>` takes any allowed extension; both stay contained under `uploads/`. `isOwnUpload()` returns the kind (`"image"` / `"file"` / `false`), and `attachmentsOk()` makes `/submit` comments and `/chat` messages carry images only in `images[]` and documents only in `files[]`, this session's uploads only, at most 5 together (`MAX_ATTACHMENTS`). A comment's `reply` gains one ` Image: <path> (<name>)` or ` File: <path> (<name>)` line per attachment, in attach order; a chat entry's lines come from `chatReplyLines()`, its images first, then its files. `sweepUploads()` bounds disk growth, and the session sweep removes the whole session a month after that. A note attachment a spec file cites is copied out first by `console.mjs keep` (`--session --upload --name --spec [--from]`): it checks both paths (the upload must sit under the session's `uploads/`, and the spec folder under `state.project`, real paths included, so a symlinked folder cannot carry the copy outside), copies with `COPYFILE_EXCL` and reuses an existing copy, prints `{ ok, path, link, image, reused }` with a forward-slash link relative to `--from`, and logs a `keep` event. It is the agent's hand, not the server's: it runs in the agent's shell only when the agent decides to cite a file, so the server still never writes into a project.
- **Answers carry their terminal token.** Each item ships `fallbackText` and `token`, so a click resolves to the literal word the prompt already expects (`approved`, `a`, `skip`). No workflow needs a second code path, and declining the console degrades to prose with nothing lost.
- **A build tells the page how long silence is normal.** The page calls a working agent stuck after two minutes without a post, which is right for a conversation and wrong for a twenty-minute build task. `agent.quietMinutes` (1 to 60, counted only while `working`) stretches the stuck, adrift and Send-grace thresholds; `quietLimitMs()` in `answers.js` is the one rule.
@@ -530,7 +534,7 @@ The bracketed flags (each entry's `bypassArgs`; a custom agent's own `bypassArgs
On a terminal it first prints a Planny banner (`renderBanner()`). With more than one on `PATH` it asks `Plan2Code opens inside an AI coding agent. Pick which one to use:` as a numbered menu, the last pick pre-selected and marked `(last used)` (Enter keeps it; a number or a CLI id also works); with one installed it uses that without asking, printing `Opening with <CLI>…` (as does a `--cli` pick); with none it exits 127 and prints `SKILL_STEPS`: how to install an agent, `cd` into the project, and type `/plan2code` (`$plan2code` in Codex). An uninstalled `--cli` pick prints that CLI's own steps after its error (its install command, `cd` into the project, then `plan2code --cli <id>` or its prompt). `--cli claude` / `devin` / `codex` (or `--cli=<name>`) skips the menu and is never forwarded; an unknown name exits 2, an uninstalled one 127. The pick is saved as `lastCli` in `~/.plan2code/launcher.json`, merged beside `lastFolder`. Before spawning, unless the forwarded args carry `--model` / `--model=` (or, for Codex, `-m`), it asks `Which model should <CLI> use?` from `models.json` plus the user's additions, Enter keeping the model the CLI last used — Ctrl+C at any launcher prompt exits 130. With no TTY on stdin there is no menu (last pick, else the first installed) and no model question: it prints that the CLI will use the model last used in it (switch it in that CLI on its own, or pass `--model <name>`). Headings, numbers, defaults, retry lines and the `plan2code:` error prefixes are colored with `paint()` (the escape values of `install.js`'s `COLORS`, copied), stdout only when stdout is a TTY and stderr only when stderr is, and never when `NO_COLOR` is set; piped runs print plain text and no banner. Other arguments are forwarded to the CLI (`plan2code --model opus`), and its exit code is passed back. `plan2code update` is the one subcommand: it is never forwarded and opens no agent, it runs `npx --yes --allow-git=all git+https://github.com/jparkerweb/plan2code.git --update`, so `install.js --update` reinstalls what `~/.plan2code/install.json` lists. On every other start, before spawning the CLI, it prints one line when a newer release is out (`Plan2Code v<latest> is out (you have v<installed>). Run: plan2code update`), compared against `~/.plan2code/bin/version.json` and read from the console's cached release check (`~/.plan2code/console/update-check.json`); it never waits on the network, and when the cache is missing or stale it refreshes it in a detached background process so the next start knows. Tests: `scripts/test-launcher.mjs` (run by `npm test`) spawns the launcher against stand-in `claude`/`devin`/`codex` commands and a temp home.
**Design constraints:**
- **Stable copy, not the package dir** — the installer normally runs from an npx cache directory npm may delete, so `installLauncher()` copies `src/launcher/plan2code.js` to `~/.plan2code/bin/plan2code.js` and the shims point there. Beside it go `update-check.mjs` (the web console's release check, for the terminal notice) and a `version.json` (the installed version). (The loop/metrics/bot wrappers point into `__dirname` and don't have this protection.)
- **Stable copy, not the package dir** — the installer normally runs from an npx cache directory npm may delete, so `installLauncher()` copies `src/launcher/plan2code.js` to `~/.plan2code/bin/plan2code.js` and the shims point there. Beside it go `update-check.mjs` (the web console's release check, for the terminal notice) and a `version.json` (the installed version). (The metrics/bot wrappers point into `__dirname` and don't have this protection.)
- **npm-style shims** — written into npm's global bin (`npm config get prefix`; `<prefix>/bin` off Windows): an sh shim everywhere, plus `.cmd` and `.ps1` on Windows, so cmd.exe, PowerShell and Git Bash all resolve it. On macOS/Linux, when the npm bin isn't writable (system Node), it falls back to `~/.local/bin`.
- **Marker-based ownership** — every shim contains `plan2code-launcher`; `uninstallLauncher()` removes only files carrying it, leaving any other `plan2code` bin alone.
- **Resolving the CLI** — the launcher walks `PATH` (+ `PATHEXT` on Windows) itself. An `.exe` is spawned directly; a `.cmd`/`.bat` wrapper goes through `cmd.exe /d /s /c` with quoted args, since Node refuses to spawn batch files without a shell. The installer warns only when neither `claude` nor `devin` is on `PATH`.
+4 -5
View File
@@ -4,13 +4,13 @@
## Code Style
- **install.js:** CommonJS, Node.js built-ins only (no external deps), ANSI colors via `COLORS` constant, readline-based prompts
- **plan2code-loop, plan2code-metrics & plan2code-bot:** TypeScript + ESM, built with tsup (target ES2022, moduleResolution: bundler)
- External deps (loop, metrics): `@inquirer/prompts`, `chalk`, `execa`, `fs-extra`, `ora`
- **plan2code-metrics & plan2code-bot:** TypeScript + ESM, built with tsup (target ES2022, moduleResolution: bundler)
- External deps (metrics): `@inquirer/prompts`, `chalk`, `execa`, `fs-extra`, `ora`
- External deps (bot): `@anthropic-ai/claude-agent-sdk`, `chalk`, `fs-extra`, `ora`
- Interactive CLI via `@inquirer/prompts` (select, input, confirm)
- **File operations:** Synchronous fs by default; the web console's fire-and-forget sweeps and `readOverview()` use `fs.promises`
- **src/web-console:** ESM, Node built-ins only, no build step. It is bundled into every skill, so it must stay dependency-free and runnable as a plain `node console.mjs`. Anything added under `src/web-console/` ships to users — put tests in `scripts/` instead.
- **US English everywhere:** docs, comments, prompts, UI text, test names and commit messages use US spelling: color, gray, behavior, center, canceled, labeled, judgment, license, defense, and -ize / -yze verbs (normalize, analyze, summarize). `scripts/test-us-english.mjs` (in `npm test`) fails on common UK spellings in tracked files. A few names keep the UK spelling because they are data contracts read in several places, and that test allow-lists them: the `greyed` field on `skills.js` entries, `{ cancelled: true }` from the folder picker and `/workspace/browse`, the `"cancelled"` event `routes.mjs` broadcasts, and the loop's `result.cancelled`. Rename one only together with every reader. `aria-labelledby` is HTML and stays; vendored files (`public/vendor/`) stay as shipped.
- **US English everywhere:** docs, comments, prompts, UI text, test names and commit messages use US spelling: color, gray, behavior, center, canceled, labeled, judgment, license, defense, and -ize / -yze verbs (normalize, analyze, summarize). `scripts/test-us-english.mjs` (in `npm test`) fails on common UK spellings in tracked files. A few names keep the UK spelling because they are data contracts read in several places, and that test allow-lists them: the `greyed` field on `skills.js` entries, `{ cancelled: true }` from the folder picker and `/workspace/browse`, the `"cancelled"` event `routes.mjs` broadcasts. Rename one only together with every reader. `aria-labelledby` is HTML and stays; vendored files (`public/vendor/`) stay as shipped.
## Gotchas / Pitfalls
@@ -19,7 +19,6 @@
- **CHANGELOG house format is not Keep a Changelog:** version headings are `## vX.Y.Z` — with a `v` prefix and **no date**. The release date is recorded only in `version.json`'s `releaseDate`. Category headings carry emoji: `### ✨ Added`, `### 🔧 Changed`, `### 🐛 Fixed`, `### 💥 Breaking`, `### 🗑️ Removed`, `### 📚 Documentation`. Older entries contain one-off variants (`🎁 Added`, `📦 Updated`, `📝 Documentation`, `🏎️ Improved`, `🧪 Testing`) — do not introduce new ones. The `.claude/skills/plan2code-changelog/` skill automates version selection and formatting; use it rather than hand-rolling an entry.
- **PowerShell mangles the CHANGELOG emoji:** `Get-Content` / `Select-String` render the `###` heading emoji as `?` under the default Windows console encoding, so a heading audit done that way reports garbage. Read `CHANGELOG.md` with a file-read or grep tool instead.
- **`.claude/skills/` is tracked:** a new repo-local skill only needs `git add` (see Architecture → Repo-Local Skills). A correction that is a workflow belongs there as a skill, linked from AGENTS.md, not as a Failure log line.
- **Loop `.gitignore` setup:** `ensureGitignore()` runs at startup in `Controller.run()` as a pre-flight step, not just inside `createTaskCommit()`. This is critical for phase mode where the Node controller doesn't handle commits — without it, `git add -A` would stage spec files.
- **Workflow file character limit:** All `src/plan2code-*.md` files (and the dashboard prompt `src/plan2code.md`) must be ≤ 20,000 characters (about 5k tokens, the usual Agent Skills guidance for a SKILL.md body). A husky pre-commit hook enforces this, and `plan2code-metrics` (`improver.ts`, `applier.ts`) applies the same number to proposed prompt edits, so change them together. History: 11,000 originally (headroom under Windsurf's old 12,000-char limit for platform-specific files, gone since 2.1.0's single-format build), 11,500 in 2.4.0, 20,000 in 2.4.1. The limit is now a prompt-size discipline only: the whole SKILL.md loads on every invocation while references load on demand, so situational detail still belongs in a reference file, and anything with one correct answer belongs in a script (see Skill scripts in the architecture doc).
- **Keep LF line endings:** `.gitattributes` pins `eol=lf` because `validate-char-count.js` counts the characters on disk, and a CRLF checkout adds about one per line to every prompt. Measure with `node scripts/validate-char-count.js` before trusting a near-limit prompt. Git never writes CRLF here, but tools do (Python writing in text mode on Windows, some editors and generators), so `scripts/fix-line-endings.mjs` puts every text file git sees back to LF: `npm run fix:eol` by hand, `--check` to only list, and it runs by itself as npm's `pretest`, in the husky pre-commit before the char count, and after every Write or Edit through the PostToolUse hook in `.claude/settings.json`. In a Python script that writes repo files, open them with `newline="\n"` (or write bytes) anyway.
- **The pre-commit hook runs only the char-count check:** `.husky/pre-commit` is `node scripts/validate-char-count.js` alone. `skills/` drift, parse errors and test failures are caught only by the tests, so run them before committing: the targeted suites for a change that stays in one area, the full `npm test` for shared code or before a PR (see "Targeted tests first" in [Development Commands](./AGENTS-development-commands.md)). Sub-package tests are separate; see [Development Commands](./AGENTS-development-commands.md).
@@ -35,5 +34,5 @@
- **The default highlight color lives in two places:** changing it means editing `DEFAULT_ACCENT` in `public/palette.js` and the light, system-dark and explicit-dark `--accent*` values in `public/styles/tokens.css`. No test ties them together.
- **Reference file sizing guideline:** Files in `src/plan2code-*-references/` directories target ~100-200 lines each (soft guideline; evaluate splitting above 300). They are NOT subject to the 20,000 character limit. The pre-commit hook (`validate-char-count.js`) only checks `src/plan2code-*.md` flat files — subdirectory contents are automatically excluded.
- **The splitting guideline has a hard ceiling — reference files cannot always be split:** each new reference costs the orchestrator a `Read references/<file>.md` line plus its fallback blockquote (~150-200 chars), which adds up across a prompt's 20,000-char budget. When a reference legitimately exceeds 300 lines (e.g. `plan2code-0-pathfinder-references/chart.md`), that is an accepted trade-off, not an oversight. Nested folders under `-references/` work: `references/` is the real path on every agent.
- **Windows global bins for loop, metrics and bot are wrappers, not `npm link`:** on win32 the installer writes `<name>.cmd` and `<name>.ps1` into `npm config get prefix`, each running `node <package>/dist/bin/<name>.js` (`install.js:1302-1323` for the loop; the metrics and bot installers do the same). After a source change, `npm run build` in the package is enough; no reinstall.
- **Windows global bins for metrics and bot are wrappers, not `npm link`:** on win32 the installer writes `<name>.cmd` and `<name>.ps1` into `npm config get prefix`, each running `node <package>/dist/bin/<name>.js` (the metrics and bot installers both do this). After a source change, `npm run build` in the package is enough; no reinstall.
- **plan2code-metrics only runs from a repo checkout:** `analyzer.ts` and `improver.ts` load `src/prompts/analyze.md` / `improve.md` at runtime through `new URL('../src/prompts/...', import.meta.url)`, resolved relative to `dist/` (`analyzer.ts:11`, `improver.ts:13`). A packed or copied `dist/` without `src/prompts/` fails.
+4 -9
View File
@@ -42,10 +42,6 @@ node scripts/capture-screenshots.mjs --only dashboard # one screen (dashboard-u
node scripts/css-snapshot.mjs --out before.json
node scripts/css-snapshot.mjs --compare before.json
# Plan2Code Loop
cd plan2code-loop && npm install # First time setup
cd plan2code-loop && npm run build # Build the CLI
# Plan2Code Metrics
cd plan2code-metrics && npm install # First time setup
cd plan2code-metrics && npm run build # Build the CLI
@@ -80,7 +76,7 @@ The full `npm test` takes minutes and drives a browser per page suite. When a ch
Run the full `npm test` when the change touches something many suites lean on: `web-console-test-helpers.mjs` or `page-test-helpers.mjs`, `looks.js` / `prefs.js` / `static.mjs` (the `SERVABLE` list), `skills.js`, `install.js`, the server's shared modules (`console.mjs`, `routes.mjs`); after a sync merge or a version bump; before opening a PR or cutting a release; and whenever you cannot say which suites a change reaches.
The root `npm test` does not run the metrics or bot vitest suites; run them in their package. `npm run dev` in loop, metrics and bot is `tsup --watch`. The web console and launcher suites isolate themselves with `PLAN2CODE_CONSOLE_HOME`, `PLAN2CODE_MODELS_DIR` and `PLAN2CODE_NO_BROWSER=1` (`scripts/web-console-test-helpers.mjs`), so they never open a browser or touch your real console home. The web console suites are split by area so `node --test` runs them in parallel. `npm test` runs them through `scripts/node-test.mjs`, which caps the files run at once at one per spare core (8 at most): each page test file drives a browser, and more at once than the cores can carry fails page tests at random (a flat 8 did on 3 and 4 core CI runners). Page tests (`test-web-console-page-*.mjs`) read styles and set scrolls only through `styleOf()`, `stylesOf()` and `scrollCard()` from `page-test-helpers.mjs`, which retry until the element is still on the page; `test-web-console-shape.mjs` fails a raw `locator.evaluate` that does either.
The root `npm test` does not run the metrics or bot vitest suites; run them in their package. `npm run dev` in metrics and bot is `tsup --watch`. The web console and launcher suites isolate themselves with `PLAN2CODE_CONSOLE_HOME`, `PLAN2CODE_MODELS_DIR` and `PLAN2CODE_NO_BROWSER=1` (`scripts/web-console-test-helpers.mjs`), so they never open a browser or touch your real console home. The web console suites are split by area so `node --test` runs them in parallel. `npm test` runs them through `scripts/node-test.mjs`, which caps the files run at once at one per spare core (8 at most): each page test file drives a browser, and more at once than the cores can carry fails page tests at random (a flat 8 did on 3 and 4 core CI runners). Page tests (`test-web-console-page-*.mjs`) read styles and set scrolls only through `styleOf()`, `stylesOf()` and `scrollCard()` from `page-test-helpers.mjs`, which retry until the element is still on the page; `test-web-console-shape.mjs` fails a raw `locator.evaluate` that does either.
## Test/Dev Env Vars
@@ -101,7 +97,7 @@ The root `npm test` does not run the metrics or bot vitest suites; run them in t
| Option | Action |
|--------|--------|
| `I` | Install the Plan2Code skills globally via the `skills` CLI, plus the global `plan2code` command (opens the dashboard in Claude Code, Devin or Codex), then print the getting-started steps. No dev tools |
| `U` | Uninstall Plan2Code: skills + `plan2code` command + `plan2code-loop` + `plan2code-metrics` + `plan2code-bot` + Claude Code status line (confirmation required) |
| `U` | Uninstall Plan2Code: skills + `plan2code` command + `plan2code-metrics` + `plan2code-bot` + Claude Code status line (confirmation required) |
| `A` | Open the ADVANCED sub-menu |
| `Q` | Quit |
@@ -110,12 +106,11 @@ The root `npm test` does not run the metrics or bot vitest suites; run them in t
| Option | Action |
|--------|--------|
| `L` | Install the skills into the current project instead of globally |
| `O` | Install plan2code-loop CLI only |
| `M` | Install plan2code-metrics CLI only |
| `S` | Install Claude Code status line only |
| `B` | Install plan2code-bot CLI only |
| `D` | Install the global `plan2code` command only |
| `E` | Everything: the skills, the `plan2code` command and every dev tool (`plan2code-loop`, `plan2code-bot`, `plan2code-metrics`, the Claude Code status line) |
| `E` | Everything: the skills, the `plan2code` command and every dev tool (`plan2code-bot`, `plan2code-metrics`, the Claude Code status line) |
| `Q` | Return to main menu |
## Non-Interactive Flags
@@ -126,7 +121,7 @@ The root `npm test` does not run the metrics or bot vitest suites; run them in t
|------|--------|
| `--build-skills` | Regenerate `skills/` from `src/`, pruning stale skills and reference files |
| `--verify-skills` | Diff local `skills/` against `src/`; exits 1 on drift. Diagnostic only, the installer rebuilds anyway |
| `--update` | No menu: reinstall what `~/.plan2code/install.json` lists (skills, launcher, statusline, loop, metrics, bot), never uninstalling; with no record, the skills and the `plan2code` command. `plan2code update` runs it through npx |
| `--update` | No menu: reinstall what `~/.plan2code/install.json` lists (skills, launcher, statusline, metrics, bot), never uninstalling; with no record, the skills and the `plan2code` command. `plan2code update` runs it through npx |
## How the Installer Works
-54
View File
@@ -1,54 +0,0 @@
# Plan2Code Loop
> Part of [AGENTS.md](../AGENTS.md) — project guidance for AI coding agents.
A separate Node.js CLI tool that autonomously implements specs by looping through tasks.
## Loop Architecture
The loop uses an **LLM-driven discovery** approach:
- Node app just orchestrates iterations and parses completion markers
- The LLM reads spec files (`overview.md`, `phase-X.md`) to discover tasks
- The LLM finds unchecked checkboxes, implements ONE task per iteration, marks it complete
- No regex parsing of markdown in Node - the AI handles all task discovery
## Loop Commands
Run `plan2code-loop` in a project (fully interactive, no flags). Build commands: see [Development Commands](./AGENTS-development-commands.md).
The CLI auto-detects specs in `./specs/`, prompts for selection if multiple found, and handles session continuation interactively. Only `specs/<feature>/overview.md` is detected; archived specs in `specs--completed/` are ignored. Session state is stored per-spec in `specs/<feature>/.plan2code-loop/`.
Exit codes: 0 all complete, 1 max iterations, 2 interrupted (SIGINT/SIGTERM saves state), 3 error. The state dir holds `config.json`, `scratchpad.md` (LLM notes), `iteration.log` (NDJSON) and `spec.hash` (a mismatch prompts "Start fresh?"). Defaults: 100 max iterations, 3 min timeout, 5 retries.
## Supported Agents
Claude Code, GitHub Copilot CLI, Devin CLI, defined in `plan2code-loop/src/agents/` (one file each, registered in `agents/index.ts`).
## Loop Modes
The CLI asks users to choose a loop mode:
- **One task per loop** (default) - Each agent invocation implements exactly one task. The Node controller handles git commits.
- **One phase per loop** - Each agent invocation implements all remaining tasks in the current phase. The LLM handles git commits (with JIRA ticket ID if provided). The controller parses multiple completion markers from a single iteration.
## Completion Markers
The LLM must output one of these formats:
- `TASK_COMPLETE: 1.1 - Task description` - Task done successfully
- `TASK_BLOCKED: 1.1 - Reason` - Cannot complete task
- `PHASE_COMPLETE` - Current phase finished (phase mode only)
- `LOOP_COMPLETE` - All phases finished
- `PREREQ_COMPLETE: P1.1 - Description` - Prerequisite verified
- `PREREQ_ASSUMED: P2.1 - Description` - Prerequisite that cannot be verified, assumed met
Parsing is in `utils/completion.ts` and is lenient: `TASK_COMPLETE: 1.1: x` and `TASK_COMPLETE[1.1]: x` also match.
## Key Source Files
| File | Purpose |
|------|---------|
| `plan2code-loop/src/controller.ts` | Main loop orchestrator |
| `plan2code-loop/src/prompt/templates.ts` | Prompt templates for both loop modes |
| `plan2code-loop/src/utils/git.ts` | `createTaskCommit()` — handles task-mode commits with footer |
| `plan2code-loop/src/cli.ts` | Interactive session setup (spec, agent, JIRA ID, loop mode, max iterations, resume/fresh) |
| `plan2code-loop/src/utils/completion.ts` | Marker parsing (`checkForCompletion` task mode, `checkForAllCompletions` phase mode) |
| `plan2code-loop/src/state/manager.ts` | Per-spec state dir, spec hash, iteration log |
| `plan2code-loop/src/agents/` | Agent adapters (claude-code, copilot-cli, devin-cli) |
+1 -1
View File
@@ -19,7 +19,7 @@ A rule stated in six files is read six times and still skipped. The console sees
- **Recommend sparingly.** "Which format survives Excel?" is the agent's to recommend; "Who is this for?" is not, because a recommendation there is an anchor people rubber-stamp.
- **A placeholder is an offer.** The Autofill button drops it into the box as a real answer, so it must be a whole answer in the person's voice and satisfy any `pattern`.
- **Titles in plain words.** Internal vocabulary (grill, frontier, fog) means nothing to a newcomer and the server refuses it. The metric words rule exists because the metrics collector scrapes `Requirements`, `Feasibility`, `Integration` and `Risk` followed by a number.
- **Loop markers refused.** `TASK_COMPLETE`, `PHASE_COMPLETE` and the rest are read by the loop CLI as completion; one in page text would end a run.
- **Completion markers refused.** `TASK_COMPLETE`, `PHASE_COMPLETE` and the rest are reserved signals; one in page text could read as a finished run.
- **The page owns its stream.** The agent never writes timestamps, versions or `agent.status` for a turn: the server stamps them, and collecting a send already sets the page to working. A second source of truth for them would drift.
- **Quoted heredoc.** Unquoted `<<EOF` lets the shell eat JSON backslashes. PowerShell has no heredoc and its pipe can mangle non-ASCII, so the patch goes through a file in the session folder, never in the project.
- **No link in the app.** Inside the desktop app the page is already in the window; printing the link sends the person to a browser tab they do not need.