loopdeck

# prompt-coach [English](README.md) | [한국어](README.ko.md) **Local-first prompt memory and coach for Claude Code and Codex.** - 🗂️ Captures every prompt you send to Claude Code / Codex into a local Markdown + SQLite archive — nothing leaves your machine. - 🧠 Scores each prompt 0–100 across five criteria and tells you which axis cost points, so you learn instead of guessing. - ✍️ Generates a copy-ready improved draft on demand (English or Korean, auto-detected) without auto-resubmitting anything. ```sh npm install -g prompt-coach prompt-coach setup --profile coach --register-mcp --open-web # then send a real Claude Code or Codex prompt and run: prompt-coach coach ``` `prompt-coach` is a developer tool that safely records prompts you enter into AI coding tools such as Claude Code and Codex, helps you find them again, analyzes weak prompting patterns, and helps you write better follow-up requests. It collects supported tool prompts locally, redacts sensitive values before storage, writes Markdown files, indexes them in SQLite, and serves a local web UI for search, review, archive scoring, prompt practice, analysis, deletion, and copy-based prompt improvement. This project is not affiliated with, endorsed by, or sponsored by Anthropic, OpenAI, or any other AI tool provider. Product names such as Claude Code and Codex are used only to describe compatibility. ## First 3-Minute Coach Loop The first success is not the web dashboard. It is seeing a score and one useful fix for a real Claude Code or Codex prompt you just sent. For most users, the happy path is: ```sh prompt-coach start --open-web prompt-coach setup --profile coach --register-mcp --open-web # send one real Claude Code or Codex coding prompt prompt-coach coach ``` Skip `--open-web` if you do not want the web workspace to open automatically on new agent sessions. Only troubleshoot after that path fails: ```sh prompt-coach doctor claude-code prompt-coach doctor codex ``` If MCP registration failed, rerun the one-command setup first: ```sh prompt-coach setup --profile coach --register-mcp --open-web ``` Manual `claude mcp add` / `codex mcp add` commands are only for advanced troubleshooting. `setup --register-mcp` is preferred because it uses the current CLI entrypoint; from a cloned checkout that means absolute Node + `dist/` paths, so Codex does not depend on `prompt-coach` being globally available in `PATH`. Open the local archive only when you want dashboard, search, history review, or export. ## Status This repository is pre-release software. - Claude Code support: MVP path - Codex support: beta adapter - Local rule-based analysis preview: implemented - Prompt Quality Score: implemented as a local deterministic `0-100` rubric - MCP prompt scoring tools: implemented as a local stdio server - Copy-based Prompt Coach: implemented, including raw-free next request briefs - Prompt Practice workspace: implemented as a local draft-and-score UI with score history and outcome feedback that do not store draft text - Transcript import: CLI only - Anonymized export: web UI and CLI preview/job flow - Benchmark v1: implemented as a local regression baseline - English/Korean web UI: implemented - External LLM analysis: no hidden provider calls from `prompt-coach`; optional MCP agent rewrite/judge packets can enter the active user-controlled Claude Code/Codex/Gemini CLI provider session when requested - Default data handling: local only ## Requirements - Node.js `>=22 <25` - pnpm `10.x` - A platform supported by `better-sqlite3` The CI target is Node 22 and Node 24. ## Quick Start There are two pieces: 1. the `prompt-coach` CLI, which owns the local server, hooks, storage, and web UI 2. the Claude Code or Codex marketplace plugin, which gives the agent an easy setup/status/open workflow The marketplace plugin does not install the CLI binary by itself. Install the CLI first, then add the marketplace. The examples below use the published CLI command `prompt-coach`. When running from a cloned development checkout, use `pnpm prompt-coach` instead. ### 1. Install The CLI After the package is published: ```sh npm install -g prompt-coach ``` For local development from this repository: ```sh git clone https://github.com/wlsdks/prompt-coach.git cd prompt-coach pnpm install # also builds dist via the prepare lifecycle pnpm setup # installs Claude Code + Codex hooks, MCP, status line, and service ``` `pnpm install` runs `pnpm build` automatically through the `prepare` lifecycle, so a fresh checkout has a working `dist/` after the install finishes. `pnpm setup` is an alias for `pnpm prompt-coach setup --profile coach --register-mcp --open-web` — one command that connects every detected agent (Claude Code and Codex), registers the MCP server with absolute paths, installs the Claude Code status line, and enables the local server on session start. ### 2. Add The Claude Code Marketplace Inside Claude Code: ```text /plugin marketplace add wlsdks/prompt-coach /plugin install prompt-coach /reload-plugins /prompt-coach:setup ``` `/prompt-coach:setup` checks that the CLI is available, previews `prompt-coach setup --profile coach --register-mcp`, asks before writing settings, and then runs the real setup if approved. ### 3. Add The Codex Marketplace From your shell: ```sh codex plugin marketplace add wlsdks/prompt-coach ``` Then run the local coach setup: ```sh prompt-coach setup --profile coach --register-mcp --open-web ``` Codex currently exposes marketplace management through `codex plugin marketplace add/upgrade/remove`. The prompt capture hook is installed by `prompt-coach setup`, which writes the Codex hook config, enables `[features].hooks`, and registers the MCP server. In a development checkout, run the same flow as `pnpm setup`; it registers MCP with absolute paths to this repo's built CLI. ### 4. Check Capture ```sh prompt-coach doctor claude-code prompt-coach doctor codex prompt-coach statusline claude-code prompt-coach buddy --once prompt-coach coach ``` Open the local archive: ```text http://127.0.0.1:17373 ``` ## Supported Platforms Release validation currently targets: - Linux x64 through GitHub Actions - Node.js 22 and 24 macOS, Linux arm64, and Windows support are intended, but they still require release smoke validation for `better-sqlite3`, filesystem permissions, and hook command behavior before a stable release claim. ## Install (Development Checkout) And Setup Options This section is for contributors and for users who want every `setup` flag documented. End users who installed `prompt-coach` from npm should follow [Quick Start](#quick-start) instead and treat this section as a reference. For local development without the agent marketplace flow: ```sh pnpm install pnpm build ``` Run the guided local coach setup: ```sh pnpm prompt-coach setup --profile coach --register-mcp ``` `setup` is intentionally explicit. Installing an npm/pnpm package should not silently edit Claude Code or Codex settings, install a login service, or start a local background server. `prompt-coach setup` is the consent step that prepares the local archive, connects supported tools that are installed on your machine, and configures the local server startup where supported. The setup command: - initializes the local data directory - detects `claude` and `codex` - installs Claude Code and/or Codex hooks for detected tools - with `--profile coach`, adds low-friction rewrite guidance through hook context instead of making you run separate score/improve commands - with `--profile coach`, installs the Claude Code status line when Claude Code is detected. Existing Claude Code status line commands are chained and restored on uninstall where possible. - with `--register-mcp`, registers the MCP server with detected Claude Code and/or Codex CLIs using the current CLI entrypoint - with `--open-web`, installs a `SessionStart` hook that ensures the local server is running and opens `http://127.0.0.1:17373` once per agent session - enables `[features].hooks` when Codex is detected - installs and starts a macOS LaunchAgent for the local server when supported - prints next steps and paths that were changed Preview setup without writing files: ```sh pnpm prompt-coach setup --profile coach --register-mcp --dry-run ``` Opt in to a Serena-like startup experience when you want the web workspace to open automatically beside Claude Code or Codex: ```sh pnpm prompt-coach setup --profile coach --register-mcp --open-web ``` This is not enabled by default. It writes an explicit `SessionStart` hook, opens the browser at most once per agent session id, and keeps the hook fail-open with no prompt body, raw path, or token output. Use passive capture only when you do not want coaching: ```sh pnpm prompt-coach setup ``` If you do not want a background service, use: ```sh pnpm prompt-coach setup --no-service pnpm prompt-coach server ``` The web UI URL is the same as in Quick Start: `http://127.0.0.1:17373`. You can still run each setup step manually. Initialize the local data directory: ```sh pnpm prompt-coach init ``` By default, data is stored under: ```text ~/.prompt-coach ``` You can use a different location with `--data-dir`: ```sh pnpm prompt-coach init --data-dir /path/to/prompt-coach-data ``` ## Start The Local Server ```sh pnpm prompt-coach server ``` The server defaults to: ```text http://127.0.0.1:17373 ``` Open that URL in a browser to use the web UI. On macOS, `setup` can install a LaunchAgent so the server starts automatically at login. You can also manage it directly: ```sh pnpm prompt-coach service install pnpm prompt-coach service status pnpm prompt-coach service start pnpm prompt-coach service stop ``` ## Connect Claude Code Install the Claude Code hook: ```sh pnpm prompt-coach install-hook claude-code ``` Optional Prompt Rewrite Guard: ```sh pnpm prompt-coach install-hook claude-code --rewrite-guard block-and-copy --rewrite-min-score 80 ``` Optional web auto-open: ```sh pnpm prompt-coach install-hook claude-code --open-web ``` `block-and-copy` uses the supported `UserPromptSubmit` decision path: weak prompts are blocked before Claude Code processes them, an improved local draft is shown, and prompt-coach tries to copy that draft to the clipboard. It does not type into the terminal, press Enter, replace the composer contents, or auto-submit anything. If the local ingest server is unavailable or ingest fails, the hook fails open and does not block the prompt. The `--rewrite-guard` flag accepts four modes: - `off` — capture only; no coaching or blocking - `context` — soft. Injects an improved draft as `additionalContext` alongside the user's submission. Claude sees both - `ask` — instructs the agent to ask one or two clarifying questions before answering. On Claude Code this uses the native `AskUserQuestion` tool; on Codex it calls the `ask_clarifying_questions` MCP tool with a native OS dialog fallback - `block-and-copy` — described above When `ask` mode triggers, prompt-coach records the event (tool, score, band, missing axes, language, prompt length) and surfaces a 7-day **Ask mode** panel on the dashboard so you can see whether the trigger gate (`length ≥ 30`, `score < 60`, not an acknowledgment) is firing on the right cases. Preview the settings change without writing: ```sh pnpm prompt-coach install-hook claude-code --dry-run ``` Diagnose the setup: ```sh pnpm prompt-coach doctor claude-code ``` `doctor` checks local server reachability, ingest token, hook installation, and MCP command access. For MCP, it first inspects known local config files and then falls back to read-only `claude mcp list` when needed. Remove the hook: ```sh pnpm prompt-coach uninstall-hook claude-code ``` The installer writes a prompt-coach command into the Claude Code settings file and creates a backup before changing an existing file. The hook command does not contain the ingest token. ## Connect Codex Beta Codex hook support is beta. Install the Codex hook: ```sh pnpm prompt-coach install-hook codex ``` Optional Prompt Rewrite Guard: ```sh pnpm prompt-coach install-hook codex --rewrite-guard block-and-copy --rewrite-min-score 80 ``` Optional web auto-open: ```sh pnpm prompt-coach install-hook codex --open-web ``` Codex support uses the same safe hook command path. Because Codex plugin-local hooks may vary by Codex version, `prompt-coach setup` / `install-hook` still writes the user-level hook config. If the local ingest server is unavailable or ingest fails, the hook fails open and does not block the prompt. Preview the `hooks.json` and `config.toml` changes without writing: ```sh pnpm prompt-coach install-hook codex --dry-run ``` Diagnose the setup: ```sh pnpm prompt-coach doctor codex ``` `doctor` checks local server reachability, ingest token, hook installation, Codex hook feature status, and MCP command access. For MCP, it first inspects known local config files and then falls back to read-only `codex mcp list` when needed. Remove the hook: ```sh pnpm prompt-coach uninstall-hook codex ``` The Codex installer targets user-level config by default: ```text ~/.codex/hooks.json ~/.codex/config.toml ``` It enables: ```toml [features] hooks = true ``` Uninstall removes the prompt-coach hook entry but leaves the Codex feature flag in place. ## Agent Wrappers Experimental `pc-claude` and `pc-codex` are experimental front-door wrappers for the initial prompt argument. They score the prompt locally, generate a redacted improvement when it is weak, and then launch the real `claude` or `codex` binary with the selected prompt. ```sh pc-claude --pc-mode auto -- "fix this" pc-codex --pc-mode auto -- "fix this" pc-codex --pc-mode auto -- exec "fix this" ``` Use dry-run first to verify what would be sent without launching the agent: ```sh pc-claude --pc-mode auto --pc-dry-run -- "fix this" pc-codex --pc-mode auto --pc-dry-run -- "fix this" ``` Wrapper options are prefixed with `--pc-*` so normal Claude/Codex options can still be forwarded. The default mode is `ask`; `--pc-mode auto` is the one-click mode that replaces a low-score initial prompt without asking. Management subcommands such as `auth`, `mcp`, `plugin`, and `login` pass through without rewriting. These wrappers do not intercept every later message typed inside an interactive session. ## Plugin Packaging This repository also ships plugin packaging artifacts: ```text .claude-plugin commands plugins/prompt-coach integrations/claude-code docs/PLUGINS.md ``` Recommended order: 1. install the `prompt-coach` CLI 2. add the agent marketplace 3. run `prompt-coach setup` or `/prompt-coach:setup` Claude Code can consume this repository as a marketplace: ```text /plugin marketplace add wlsdks/prompt-coach /plugin install prompt-coach /reload-plugins /prompt-coach:setup ``` The Claude Code plugin provides slash commands: ```text /prompt-coach:setup /prompt-coach:status /prompt-coach:guard /prompt-coach:buddy /prompt-coach:coach /prompt-coach:score /prompt-coach:judge /prompt-coach:improve-last /prompt-coach:habits /prompt-coach:open ``` `/prompt-coach:guard` opens an interactive picker (off / context / ask / block-and-copy) that flips the `UserPromptSubmit` rewrite-guard mode without requiring you to remember CLI flags. Run `prompt-coach hook status` to see the mode currently installed for each detected tool. `/prompt-coach:setup` runs `prompt-coach setup --dry-run` first, asks before writing local settings, and can optionally install a small Claude Code `statusLine` indicator with the latest prompt score: ```sh pnpm prompt-coach install-statusline claude-code ``` If another Claude Code HUD is already installed, prompt-coach preserves it by running both commands through one chained `statusLine` command. Uninstalling prompt-coach restores the previous command when it was captured during install. For Claude Code or Codex, open a second terminal pane beside the agent and run the always-on prompt buddy: ```sh pnpm prompt-coach buddy ``` Use `pnpm prompt-coach buddy --once` for a one-shot text snapshot, or `pnpm prompt-coach buddy --json` for automation. The Codex package under `plugins/prompt-coach` contains a `.codex-plugin` manifest, a fail-open `UserPromptSubmit` hook, and a small skill that helps Codex install, diagnose, and use the local archive. Claude Code prompt capture is exposed through its documented hook settings, so `integrations/claude-code/settings.example.json` is provided as a manual example. For normal use, prefer: ```sh pnpm prompt-coach setup ``` The explicit setup command is still required because plugin discovery should not silently edit user settings, install a login service, or start a local server. See `docs/PLUGINS.md` for the packaging boundary and manual configuration notes. Render the Claude Code status line manually: ```sh pnpm prompt-coach statusline claude-code ``` Render a side-pane buddy snapshot manually: ```sh pnpm prompt-coach buddy --once ``` Codex can add the same repository as a marketplace: ```sh codex plugin marketplace add wlsdks/prompt-coach ``` After that, use `prompt-coach setup` to install the Codex hook and enable Codex hooks. ## CLI List prompts: ```sh pnpm prompt-coach list ``` Search prompts: ```sh pnpm prompt-coach search "migration plan" ``` Show a prompt Markdown body: ```sh pnpm prompt-coach show <prompt-id> ``` Delete a prompt: ```sh pnpm prompt-coach delete <prompt-id> ``` Open a prompt in the local web UI: ```sh pnpm prompt-coach open <prompt-id> ``` Rebuild SQLite/FTS from Markdown: ```sh pnpm prompt-coach rebuild-index ``` Preview and import JSONL transcripts: ```sh pnpm prompt-coach import --dry-run --file ./transcript.jsonl --save-job pnpm prompt-coach import --execute --file ./transcript.jsonl pnpm prompt-coach import-job <job-id> ``` Import is currently CLI-centered. The web UI can browse imported prompts through the normal archive and imported-only filters, but there is no web import upload screen. Create and execute an anonymized export: ```sh pnpm prompt-coach export --anonymized --preview --preset anonymized_review --json pnpm prompt-coach export --anonymized --job <export-job-id> --json ``` The web UI exposes only anonymized export. Raw export is not implemented. Previewed export jobs expire and are invalidated when the selected prompt set, project policy versions, redaction version, or preview counts change. Generate a copy-based Prompt Coach draft: ```sh pnpm prompt-coach coach pnpm prompt-coach coach --json pnpm prompt-coach improve --text "make this request clearer" --json pnpm prompt-coach improve --latest --json ``` Score accumulated prompt habits without returning prompt bodies: ```sh pnpm prompt-coach score --json pnpm prompt-coach score --latest --json pnpm prompt-coach score --tool codex --json ``` ## Local Analysis Preview Prompt detail views include a local rule-based analysis preview. It summarizes whether a prompt includes clear targets, context, constraints, output format, and verification criteria. Each prompt also receives a deterministic `0-100` Prompt Quality Score with a checklist-based breakdown. This preview runs locally against the stored, redacted prompt body. It does not call an external LLM provider. ## Project Instruction Review The Projects screen can analyze project-local `AGENTS.md` and `CLAUDE.md` files. The review stores a local snapshot with file names, hashes, timestamps, checklist status, score, and improvement hints. It does not store or return instruction file bodies, raw absolute paths, or external LLM results. The score is a deterministic local rubric for project context, agent workflow, verification commands, privacy/safety, and reporting rules. ## MCP Prompt Scoring `prompt-coach` can expose the same local Prompt Quality Score to Claude Code, Codex, or any MCP client through a stdio MCP server: ```sh prompt-coach mcp ``` The MCP server exposes thirteen tools: - `get_prompt_coach_status`: check whether the local archive is initialized, whether prompts have been captured, and which MCP tool to call next. - `coach_prompt`: run the default one-call agent workflow for Claude Code or Codex: local readiness, latest prompt score, approval-required rewrite, recent habit review, project instruction review, and next request guidance. - `score_prompt`: score either direct prompt text, a stored `prompt_id`, or the latest stored prompt. - `improve_prompt`: generate an approval-ready improved prompt draft for direct prompt text, a stored `prompt_id`, or the latest stored prompt. The result also includes a `clarifying_questions` array (with JSON-Schema-shaped `answer_schema.examples`) the agent should ask the user via its native ask UI. - `apply_clarifications`: take the user's verbatim answers (each must be tagged `origin: "user"`) and compose the final approval-ready draft. Use this after the agent has collected answers through its own ask UI. - `ask_clarifying_questions`: prompt-coach drives the entire ask-then-apply flow itself. Three layered paths, in order: 1. MCP `elicitation/create` when the client advertises `capabilities.elicitation` (Claude Code 2.1.76+). 2. Native OS dialog (macOS `osascript`, Linux `zenity`, Windows PowerShell `Microsoft.VisualBasic.InputBox`) when the caller opts in via `allow_native_dialog: true` or `PROMPT_COACH_NATIVE_DIALOG=1`. Useful on Codex today, before `ask_user_question` ships upstream. 3. Otherwise returns `clarifying_questions` metadata (`interaction_status: unsupported|declined|timeout`). Never auto-submits a rewrite. - `record_clarifications`: persist the user's verbatim answers and the resulting draft against a stored prompt in the local archive (`prompt_improvement_drafts`). Returns metadata only (`draft_id`, `answers_count`, `changed_sections`, …) — the prompt body and the draft text are never echoed in the response. Local-only write tool. - `prepare_agent_rewrite`: prepare one locally redacted prompt packet, local score metadata, local baseline draft, and rewrite contract so the active Claude Code/Codex/Gemini CLI session can semantically improve the prompt. - `record_agent_rewrite`: save that agent-produced rewrite as a redacted improvement draft after user approval, without returning the rewrite body. - `score_prompt_archive`: score accumulated prompt habits across recent stored prompts and return aggregate score, recurring gaps, a practice plan, a next prompt template, and low-score prompt ids. - `review_project_instructions`: review local `AGENTS.md` / `CLAUDE.md` instruction files for the latest or selected project and return score, checklist status, and improvement hints. - `prepare_agent_judge_batch`: prepare a bounded, locally redacted prompt packet and rubric for the active Claude Code/Codex/Gemini CLI session to judge. `prompt-coach` does not call the provider for you. - `record_agent_judgments`: store advisory scores and notes produced by the active agent session, without storing prompt bodies or raw paths. All read tools are local-only and declare an MCP `outputSchema` for structured JSON metadata plus a text JSON fallback. `record_agent_rewrite` and `record_agent_judgments` are non-destructive write tools. Archive-backed local tools do not return stored prompt bodies, raw absolute paths, secrets, or hidden external LLM results. Agent rewrite/judge modes are opt-in and use the current agent session as the rewriter or evaluator. Practical agent prompts: ```text Use prompt-coach coach_prompt and give me the one-call coaching result for my latest request. Do not auto-submit the rewrite. Use prompt-coach get_prompt_coach_status and tell me whether prompt capture is working before you score anything. Use prompt-coach score_prompt with latest=true and tell me what to improve in my last request. Use prompt-coach improve_prompt with latest=true and give me an approval-ready draft I can copy and resubmit. Use prompt-coach ask_clarifying_questions with prompt: "<my draft>". If your client supports MCP elicitation, prompt-coach will ask me via your native ask UI; otherwise return the clarifying_questions metadata so you can ask through AskUserQuestion or Codex ask_user_question and pass my answers to apply_clarifications. Use prompt-coach prepare_agent_rewrite with latest=true. Rewrite that redacted prompt yourself, ask for my approval, then call record_agent_rewrite if I want the draft saved. Use prompt-coach score_prompt_archive for recent Codex prompts and summarize my top recurring prompt habit gaps. Use prompt-coach review_project_instructions with latest=true and tell me whether my AGENTS.md/CLAUDE.md rules are strong enough for coding agents. Use prompt-coach prepare_agent_judge_batch with selection=low_score and max_prompts=5. Judge those redacted prompts yourself, then call record_agent_judgments with your scores and suggestions. ``` The tools return score metadata, checklist breakdowns, warnings, recurring gaps, approval-ready rewrite drafts, and improvement hints. They do not store direct prompt text or make hidden external LLM calls. Archive-backed score/rewrite flows do not return stored original prompt bodies. The archive scoring tool also avoids raw absolute paths. The project instruction review tool also avoids instruction file bodies and raw absolute paths. The status tool returns only safe counts, latest prompt metadata, available tool names, and next actions. Agent-judge packets are different: when explicitly requested, they return locally redacted prompt bodies so the active Claude Code/Codex/Gemini CLI session can judge them. This is documented in [Legal usage guide](docs/LEGAL_USAGE_GUIDE.md). `prompt-coach` does not extract or proxy Claude.ai OAuth tokens, Claude Code internal auth tokens, OpenAI/Codex/ChatGPT session tokens, or provider API keys. Example Claude Code registration: ```sh claude mcp add --transport stdio prompt-coach -- prompt-coach mcp ``` Example Codex registration: ```sh codex mcp add prompt-coach -- prompt-coach mcp ``` Those manual examples assume the published `prompt-coach` binary is available in `PATH`. For local development, prefer: ```sh pnpm setup ``` or rerun: ```sh pnpm prompt-coach setup --profile coach --register-mcp --open-web ``` The setup command registers MCP with absolute Node + `dist/cli/index.js` paths, which is the safer Codex configuration for a cloned checkout. If you use a custom data directory: ```sh prompt-coach mcp --data-dir /path/to/prompt-coach-data ``` ## Benchmark Benchmark v1 measures local regression signals for privacy, retrieval, rule-based prompt improvement, `coach_prompt` actionability, prompt quality score calibration, analytics, and latency: ```sh pnpm benchmark pnpm benchmark -- --json ``` The benchmark uses synthetic fixtures only. It is a local baseline, not a claim that real user prompt quality is fully solved. ## Release Smoke Run the local release smoke before publishing or tagging a beta: ```sh pnpm smoke:release ``` The smoke script builds the package, creates an isolated temporary data directory and HOME, starts the local server, captures fixture-like Claude Code and Codex prompts, verifies CLI list/search/show/delete/rebuild-index, checks SQLite WAL/FTS5, and confirms deleted prompt metadata is removed. Browser regression smoke is also available: ```sh pnpm e2e:browser ``` It checks the archive, prompt detail, Prompt Coach copy/save flow, projects, anonymized export, and mobile overflow against a real local server. ## Storage `prompt-coach` treats Markdown as the source of truth and SQLite as an index. Default files: ```text ~/.prompt-coach/config.json ~/.prompt-coach/hook-auth.json ~/.prompt-coach/prompt-coach.sqlite ~/.prompt-coach/prompts/ ~/.prompt-coach/logs/ ~/.prompt-coach/quarantine/ ~/.prompt-coach/spool/ ``` On POSIX systems, prompt-coach creates sensitive directories as `0700` and token/config files as `0600`. ## Privacy And Security Default behavior: - Prompt capture is local to `127.0.0.1`. - Hook ingest uses a local bearer token stored in `hook-auth.json`. - The browser UI uses a same-origin session cookie and CSRF token. - Sensitive values are redacted before Markdown, SQLite, and FTS indexing in `mask` mode. - External LLM analysis is never triggered as a hidden background call by `prompt-coach`. Optional MCP agent rewrite/judge workflows can return redacted prompt packets to the active user-controlled Claude Code, Codex, or Gemini CLI session when requested, and that agent may send the packet through its provider session according to the user's tool setup. - Prompt Coach is copy-based. It does not automatically type into, replace, or resubmit prompts into Claude Code or Codex. - Prompt Rewrite Guard is opt-in. In `block-and-copy` mode it blocks weak prompts and offers a copied local rewrite for manual paste/enter. In `context` mode it adds model-visible rewrite guidance but does not replace the original prompt. - Settings and local diagnostics may show local filesystem paths to the local user. Browser prompt/archive/export surfaces mask prompt-body paths and avoid raw prompt identifiers. Important limits: - This tool stores prompts you submit to connected tools. Only enable hooks where you are allowed to store that content. - Redaction is best-effort and should not be treated as a complete data loss prevention system. - Deletion removes prompt-coach Markdown and SQLite rows, but it does not erase copies that may exist in terminal history, editor buffers, backups, filesystem snapshots, or the upstream AI tool transcript. - This project does not extract, store, proxy, sell, or reuse Claude.ai OAuth tokens, Claude Code internal auth tokens, OpenAI/Codex session tokens, or ChatGPT account tokens. ## Remove Data Remove a single prompt: ```sh pnpm prompt-coach delete <prompt-id> ``` Remove hooks: ```sh pnpm prompt-coach uninstall-hook claude-code pnpm prompt-coach uninstall-hook codex ``` Remove all prompt-coach data: ```sh rm -rf ~/.prompt-coach ``` Use your configured `--data-dir` path if you initialized prompt-coach somewhere else. ## Development Run the full local gate: ```sh pnpm format pnpm test pnpm lint pnpm build pnpm pack:dry-run ``` The dry-run package should include built CLI files, built web assets, README, and release documentation. See [Package contents](docs/PACKAGE_CONTENTS.md) before publishing to confirm which files ship to npm, and [Pre-publish privacy audit](docs/PRE_PUBLISH_PRIVACY_AUDIT.md) for the current privacy review checklist. ## Contributing Please read [CONTRIBUTING](CONTRIBUTING.md), [CODE OF CONDUCT](CODE_OF_CONDUCT.md), [SUPPORT](SUPPORT.md), and [SECURITY](SECURITY.md) before opening issues, pull requests, or security reports. ## Documentation - [PRD](docs/PRD.md) - [Phase 2 PRD](docs/PRD_PHASE2.md) - [Package contents](docs/PACKAGE_CONTENTS.md) - [Pre-publish privacy audit](docs/PRE_PUBLISH_PRIVACY_AUDIT.md) - [Efficiency review](docs/EFFICIENCY_REVIEW.md) - [Architecture](docs/ARCHITECTURE.md) - [Tech spec](docs/TECH_SPEC.md) - [Implementation plan](docs/IMPLEMENTATION_PLAN.md) - [Adapter guide](docs/ADAPTERS.md) - [Release checklist](docs/RELEASE_CHECKLIST.md) - [Security policy](SECURITY.md) ## License MIT