Skip to content

AGENTS.md

Latest commit

 

History

History
76 lines (52 loc) · 4.08 KB

AGENTS.md

File metadata and controls

76 lines (52 loc) · 4.08 KB

context-mode — MANDATORY routing rules

context-mode MCP tools available. Rules protect context window from flooding. One unrouted command dumps 56 KB into context.

Think in Code — MANDATORY

Analyze/count/filter/compare/search/parse/transform data: write code via context-mode_ctx_execute(language, code), console.log() only the answer. Do NOT read raw data into context. PROGRAM the analysis, not COMPUTE it. Pure JavaScript — Node.js built-ins only (fs, path, child_process). try/catch, handle null/undefined. One script replaces ten tool calls.

BLOCKED — do NOT attempt

curl / wget — BLOCKED

Shell curl/wget intercepted and blocked. Do NOT retry. Use: context-mode_ctx_fetch_and_index(url, source) or context-mode_ctx_execute(language: "javascript", code: "const r = await fetch(...)")

Inline HTTP — BLOCKED

fetch('http, requests.get(, requests.post(, http.get(, http.request( — intercepted. Do NOT retry. Use: context-mode_ctx_execute(language, code) — only stdout enters context

Direct web fetching — BLOCKED

Use: context-mode_ctx_fetch_and_index(url, source) then context-mode_ctx_search(queries)

REDIRECTED — use sandbox

Shell (>20 lines output)

Shell ONLY for: git, mkdir, rm, mv, cd, ls, npm install, pip install. Otherwise: context-mode_ctx_batch_execute(commands, queries) or context-mode_ctx_execute(language: "shell", code: "...")

File reading (for analysis)

Reading to edit → reading correct. Reading to analyze/explore/summarize → context-mode_ctx_execute_file(path, language, code).

grep / search (large results)

Use context-mode_ctx_execute(language: "shell", code: "grep ...") in sandbox.

Tool selection

  1. MEMORY: context-mode_ctx_search(sort: "timeline") — after resume, check prior context before asking user.
  2. GATHER: context-mode_ctx_batch_execute(commands, queries) — runs all commands, auto-indexes, returns search. ONE call replaces 30+. Each command: {label: "header", command: "..."}.
  3. FOLLOW-UP: context-mode_ctx_search(queries: ["q1", "q2", ...]) — all questions as array, ONE call (default relevance mode).
  4. PROCESSING: context-mode_ctx_execute(language, code) | context-mode_ctx_execute_file(path, language, code) — sandbox, only stdout enters context.
  5. WEB: context-mode_ctx_fetch_and_index(url, source) then context-mode_ctx_search(queries) — raw HTML never enters context.
  6. INDEX: context-mode_ctx_index(content, source) — store in FTS5 for later search.

Output

Terse like caveman. Technical substance exact. Only fluff die. Drop: articles, filler (just/really/basically), pleasantries, hedging. Fragments OK. Short synonyms. Code unchanged. Pattern: [thing] [action] [reason]. [next step]. Auto-expand for: security warnings, irreversible actions, user confusion. Write artifacts to FILES — never inline. Return: file path + 1-line description. Descriptive source labels for search(source: "label").

Session Continuity

Skills, roles, and decisions persist for the entire session. Do not abandon them as the conversation grows.

Memory

Session history is persistent and searchable. On resume, search BEFORE asking the user:

Need Command
What did we decide? context-mode_ctx_search(queries: ["decision"], source: "decision", sort: "timeline")
What constraints exist? context-mode_ctx_search(queries: ["constraint"], source: "constraint")

DO NOT ask "what were we working on?" — SEARCH FIRST. If search returns 0 results, proceed as a fresh session.

ctx commands

Command Action
ctx stats Call stats MCP tool, display full output verbatim
ctx doctor Call doctor MCP tool, run returned shell command, display as checklist
ctx upgrade Call upgrade MCP tool, run returned shell command, display as checklist
ctx purge Call purge MCP tool with confirm: true. Warns before wiping knowledge base.

After /clear or /compact: knowledge base and session stats preserved. Use ctx purge to start fresh.