Skip to main content
Glama
AGENT_PROMPT.md3.36 kB
# Docdex — Agent Prompt You are interacting with Docdex, a local documentation indexer/search daemon. It keeps a per-repo index of Markdown/text files on disk and serves search/snippet results over HTTP and CLI. No external services are used; all data stays local. ## What Docdex does - Builds a tantivy-based index under `<repo>/.docdex/index`. - Serves HTTP endpoints: `/search`, `/snippet`, `/healthz`. - CLI commands mirror the API: `docdexd index|serve|query|ingest|self-check`. - Optional MCP mode (stdio) exposes `docdex.search` and `docdex.index` tools for MCP-aware clients. - Watches files while serving to keep the index fresh. ## How to use - Install via npm: `npm i -g docdex` (or `npx docdex --version`). - Build an index: `docdexd index --repo /path/to/repo`. - Serve API: `docdexd serve --repo /path/to/repo --host 127.0.0.1 --port 46137 --log info --auth-token <token>` (or add `--secure-mode=false` for token-free local use). - Query via CLI: `docdexd query --repo /path/to/repo --query "term" --limit 5`. - Health: `GET /healthz` should return `ok`. ## Security/constraints - Defaults bind to `127.0.0.1`; secure mode is on by default and requires an auth token, loopback-only allowlist, and default rate limiting. Add `--secure-mode=false` (and set `--allow-ip`/`--rate-limit-per-min`) when you need broader access. Respect local-only behavior unless configured otherwise. - Index stays on disk; do not upload corpus or snippets externally. - Respect rate limits and request size defaults (`max_query_bytes`, `max_request_bytes`, `max_limit`). ## Paths and binaries - Binary name: `docdexd` (`docdex` alias via npm). - Index path: `<repo>/.docdex/index` (or `DOCDEX_STATE_DIR`). - Supported platforms: macOS (arm64/x64), Linux glibc (arm64/x64), Linux musl (arm64/x64), Windows (x64/arm64) with matching release artifacts. ## Environment overrides (common) - `DOCDEX_STATE_DIR` — override index location. - `DOCDEX_DOWNLOAD_REPO` — owner/repo for release assets (npm installer). - `DOCDEX_LIBC` — force `gnu` or `musl` on Linux. - `DOCDEX_GITHUB_TOKEN` — authenticated downloads of release assets (avoids rate limits/private releases). ## Agent guidance - Keep queries concise; prefer summary-only when possible (`snippets=false`), then fetch snippets for selected docs. - Avoid sending sensitive content elsewhere; Docdex is local—keep data local. - When returning snippets, include `rel_path` so humans can navigate the source. - If your client supports MCP, use `docdex.search` (concise queries, low `limit`) for repo-specific context, `docdex.index` when results look stale, `docdex.files` to list indexed docs, and `docdex.stats` to confirm doc counts/recency; otherwise keep using HTTP/CLI. - When integrating Docdex into a new repo, ensure `.docdex/` (especially `.docdex/index/`) is listed in `.gitignore` so index artifacts are never committed. - MCP client setup: register a server named `docdex` that runs `docdexd mcp --repo <repo> --log warn --max-results 8` (or `docdex mcp ...` from npm). Use the Docdex helper to auto-add wherever possible: `docdex mcp-add --repo <repo>` (add `--all` to attempt every supported client, or `--remove` to uninstall). Direct Codex CLI command: `codex mcp add docdex -- docdexd mcp --repo <repo> --log warn --max-results 8`. Then use the MCP tools instead of shelling out.

MCP directory API

We provide all the information about MCP servers via our MCP API.

curl -X GET 'https://glama.ai/api/mcp/v1/servers/bekirdag/docdex'

If you have feedback or need assistance with the MCP directory API, please join our Discord server