القائمة
Ground documentation answers in the files that support them. Use when answering questions from a documentation tree or reporting related source paths.
التثبيت باستخدام Codex أو Claude انسخ هذا Prompt والصقه في Codex أو Claude أو مساعد آخر ليراجع صفحة Skill ويثبّتها لك.
استنادا إلى تصنيف SOC المهني
| name | source-citation |
| description | Ground documentation answers in the files that support them. Use when answering questions from a documentation tree or reporting related source paths. |
Search before answering, then read the smallest set of relevant files needed to verify the claim.
Return paths relative to the documentation root in related_documents. Include only files that directly support the response, and do not invent a path when the documentation does not answer the question.
Build, configure, run, test, and package manifest-defined huglets. Use when creating or changing an agent crate with huggr.toml, SYSTEM.md, a Rust response contract, tool grants, context policy, runtime arguments, traces, or a standalone CLI/MCP/Python artifact.
Define and run huglets in Python with typed contract objects, sync or async callable tools, streaming events, feedback, stats, and standard traces. Use for the huggr-agents Python runtime API, or when deciding between a Python-defined agent and a generated Python wheel from a Rust-defined agent.
Define and run huglets in TypeScript or JavaScript through the huggr-agents WASM runtime, with Node or browser storage, callable tools, streaming events, feedback, and cross-language trace verification. Use when building a Node agent, a browser-hosted agent, or a custom TypeScript host around huggr-core.
Build or modify a Chrome Manifest V3 extension that hosts the Huggr WASM brain, maps browser capability schemas to chrome.* implementations, persists settings and traces in IndexedDB, and calls an OpenAI-compatible model. Use for examples/chrome-extension, custom browser-agent UI, capability dispatch, extension packaging, or WASM/Chrome debugging.
Inspect, replay, verify, compare, and analyze Huggr traces, lineage, costs, tool activity, feedback, and offline insights. Use when an ask fails, a trace no longer verifies, behavior changes after a reducer or policy edit, costs regress, a resumed session looks wrong, or an agent needs trace-driven improvement suggestions.
