Files
my-pi/pi-web-search
云服务部-叶林立andDevin <158243242+devin-ai-integration[bot]@users.noreply.github.com> 0c45b1a438 docs: 拆分 AGENTS.md,根文件只保留全局规则
根 AGENTS.md 从约 78KB 压缩到约 10KB,避免作为常驻规则加载时被截断:
只保留仓库定位、扩展清单、跨扩展硬规则与通用验证命令。各扩展的维护
约定与验证命令移入对应目录的 AGENTS.md(新增 extensions/AGENTS.md 承载
系统提示词与根目录 TS/LSP 约定),上游来源记录统一放进各目录的中文
UPSTREAM.md,并补齐 pi-permission-system 与 pi-rtk-optimizer 的来源记录。

Generated with [Devin](https://devin.ai)

Co-Authored-By: Devin <158243242+devin-ai-integration[bot]@users.noreply.github.com>
2026-10-20 19:43:22 -07:00
..

pi-web-search

Unified web search and fetch for Pi. The main model only ever sees two tools — web_search and web_fetch — provider selection happens inside the extension.

  • web_search(query, max_results?, site?, published_after?, published_before?, depth?) routes each query to Tavily, Exa, or Keenable using the session's current main model (ctx.model + ctx.modelRegistry.complete), guided by per-provider rules. Results include details.provider / details.routedBy for debugging.
  • web_fetch(urls, max_characters?) always reads pages via Exa (web_fetch_exa), degrading silently to Keenable then Tavily.

Providers (clients/)

Plain fetch calls — no MCP adapter, no upstream extension wrappers:

  • clients/tavily.ts — POST api.tavily.com/search|extract, Bearer key. Requires a tavily credential.
  • clients/keenable.ts — api.keenable.ai, keyless /public endpoints by default, a keenable credential raises rate limits. Adapted from @keenable/pi-search.
  • clients/exa.ts — JSON-RPC over mcp.exa.ai/mcp (Streamable HTTP): initialize once, cache mcp-session-id, tools/call per request, re-init on session expiry. An exa credential is optional (x-api-key header).

Keys live in Pi's auth.json as one api_key credential per provider (tavily, exa, keenable), read by auth.ts with upstream resolution semantics — literal values, $VAR/${VAR} interpolation, and !command indirection (e.g. "!op read op://Vault/Tavily/key"). configure.sh search writes them (atomically, mode 600, preserving all other entries); process environment variables are not read directly — use "key": "$TAVILY_API_KEY" to indirect through env. KEENABLE_API_URL stays env-only, or can be stored in the keenable credential's env map.

Routing

router.ts asks the current model to answer with exactly one provider name, given a prompt containing the per-provider strengths and the query's requested filters. Decisions are cached per session (LRU, 32 entries). Failure paths are deterministic: an unroutable answer or a dead provider falls back across the remaining available providers — never a heuristic choice over the model.