What is michat-skill-implementation?

It is the repo-guided process for implementing a new MiChat agent skill and wiring its tools, following the A–E steps and profile-based visibility rules.

Do I need to modify every profile?

Skills should be added to the relevant profile(s) only, and only profiles that enable the corresponding toolset will see/use the tool.

How do I verify setup and behavior?

Use the exact verification commands: python -m michat.shell --profile --debug and python -m michat.chat --profile --debug, then check the tool list, skill list, and expected behavior as documented in the steps.

michat-skill-implementation

npx machina-cli add skill filmicgaze/MiChat-desktop/michat-skill-implementation --openclaw

Files (1)

SKILL.md

2.4 KB

Architecture assumptions

MiChat is profile-driven: profiles/<profile>/ holds profile.json, system_prompt.txt, skills/, ui.json.
Agent skills are SKILL.md files under profiles/<profile>/skills/<slug>/SKILL.md.
Tool capability boundaries are enforced by toolsets + enabled_toolsets in profile.json.
MiChat requires a profile; no legacy root ./skills fallback exists.

Implementation loop

Follow the steps in order. Do not improvise or skip steps. If anything is unclear or missing, ask for clarification before editing.

A) Decide placement

Decide if this is a new toolset or belongs to an existing toolset.
Choose which profiles should get it (toolset enablement + skill file).

B) Implement code

Add/update the Python tool implementation.
Register the tool in the correct toolset (toolsets registry).
Ensure only profiles that enable the toolset can see/use the tool.

C) Add MiChat agent skill files

Create/edit SKILL.md under the relevant profile(s) only.
YAML frontmatter: require only name and description.
If allowed_tools is used (per spec), list only tool names that exist in enabled toolsets for that profile.

D) Wire and sanity check

Ensure manifest injection lists available skills for the active profile.
Ensure read_skill reads from the selected profile skills dir only.

E) Output verification steps (for Nicholas)

Provide exact commands, e.g.:
- python -m michat.shell --profile <name> --debug
- python -m michat.chat --profile <name> --debug
State exactly what to check (tool list, skill list, expected behavior).
Use the repo’s current run entrypoints; the examples above are current as of this repo state.

House rules / constraints

Follow the steps above as written; do not improvise or skip steps.
Ask for clarification when information is missing or ambiguous.
Never create/copy secrets or tokens; never commit token.json or client_secret*.json.
Do not use AppData or write outside the repo/exe folder; runtime state belongs under ./data and profile resources under profiles/<profile>/...
Keep changes small and reversible; avoid refactors unless asked.
If behavior depends on runtime UI, ask Nicholas to run and paste logs instead of guessing.

Source

git clone https://github.com/filmicgaze/MiChat-desktop/blob/main/.codex/skills/michat-skill-implementation/SKILL.mdView on GitHub

Overview

This skill provides a structured workflow to implement MiChat agent skills and wire tools across profiles. It explains architecture assumptions like profiles/<profile>/ and how to place SKILL.md files. It also outlines a repeatable implementation loop (A–E) to ensure proper tool integration, visibility, and verification.

How This Skill Works

Follow the documented implementation loop: A) decide placement (new toolset vs existing), B) implement code (Python tool, toolset registry), C) add/edit SKILL.md under the relevant profile with proper allowed_tools, D) wire and perform sanity checks (manifest injection, read_skill scope), and E) run output verification steps. This ensures only enabled profiles see the tool and that skills are discoverable from the active profile.

When to Use It

When adding a new toolset or integrating a new tool into MiChat.
When associating a skill with an existing toolset and ensuring correct registration.
When deploying to multiple profiles and enforcing per-profile enablement of toolsets.
When validating that SKILL.md lists only tools from enabled toolsets for the profile.
When performing the prescribed verification steps with the recommended commands.

Quick Start

Step 1: Decide placement (new toolset vs existing) and choose target profiles to enable it.
Step 2: Implement the Python tool, register it in the toolset, and create/edit the profile's SKILL.md with proper frontmatter.
Step 3: Run verification: python -m michat.shell --profile <name> --debug and python -m michat.chat --profile <name> --debug; check tool list, skill list, and expected behavior.

Best Practices

Follow the A–E steps strictly in order; do not improvise.
Keep YAML frontmatter minimal in SKILL.md (name and description only).
If using allowed_tools, list only tools that exist in enabled toolsets for the profile.
Test changes with profile-specific runs: python -m michat.shell --profile <name> --debug and python -m michat.chat --profile <name> --debug.
Keep changes small and reversible; avoid broad refactors unless requested.

Example Use Cases

Add a new toolset for a custom data-fetching tool and create a corresponding michat-skill-implementation SKILL.md under profiles/sales/skills/forecast/.
Wire an existing weather data tool into a new MiChat skill in profiles/marketing, ensuring the tool is exposed only to profiles with the toolset enabled.
Register the Python tool in the correct toolset registry and update enabled_toolsets in profile.json.
Verify that manifest injection lists the new skill for the active profile and that read_skill reads only from profiles/<profile>/skills.
Run python -m michat.shell --profile <name> --debug followed by python -m michat.chat --profile <name> --debug to confirm tool and skill visibility and expected behavior.

Frequently Asked Questions

Add this skill to your agents