context-mode

Raw tool output floods your context window. Use context-mode MCP tools to keep raw data in the sandbox.

Tool Selection

1. GATHER: batch_execute(commands, queries) — Primary tool for research. Runs all commands, auto-indexes, and searches. ONE call replaces many individual steps.
2. FOLLOW-UP: search(queries: ["q1", "q2", ...]) — Use for all follow-up questions. ONE call, many queries.
3. PROCESSING: execute(language, code) or execute_file(path, language, code) — Use for API calls, log analysis, and data processing.
4. WEB: fetch_and_index(url) then search(queries) — Fetch, index, then query. Never dump raw HTML.

Rules

• DO NOT use Bash for commands producing >20 lines of output — use execute or batch_execute.
• DO NOT use Read for analysis — use execute_file. Read IS correct for files you intend to Edit.
• DO NOT use WebFetch — use fetch_and_index instead.
• DO NOT use curl/wget in Bash — use execute or fetch_and_index.
• Bash is ONLY for git, mkdir, rm, mv, navigation, and short commands.

Output

• Keep responses under 500 words.
• Write artifacts (code, configs) to FILES — never return them as inline text.
• Return only: file path + 1-line description.