Tatsh

Tatsh/wiswa-mcp

Agent Claude Code

Synchronises the badge list in docs/badges.rst with README.md. Use after editing the README badge block or when docs and README drift.

Tatsh/wiswa-mcp

Agent Claude Code

Updates CHANGELOG.md with Keep-a-Changelog entries derived from commits and diffs since the last release. Use before a version bump or when the user asks for changelog work.

Tatsh/wiswa-mcp

Agent Claude Code

Audits Click CLI commands for project conventions (option naming, help-string punctuation, path types) and applies fixes. Use after adding or changing Click commands.

Tatsh/wiswa-mcp

Agent Claude Code

Fixes prose style, grammar, spelling, and punctuation in comments, docstrings, and user-facing string literals. Touches no code logic or identifiers. Use for documentation polish passes.

Tatsh/wiswa-mcp

Agent Claude Code

Identifies uncovered code paths and writes targeted pytest tests to close the gap. Use when test coverage falls below the project target or before a release.

Tatsh/wiswa-mcp

Agent Claude Code

Audits public APIs for missing or incomplete NumPy-style docstrings and writes them. Use after adding new public functions, classes, or modules.

Tatsh/wiswa-mcp

Agent Claude Code

Fixes markdownlint-cli2 errors in Markdown and MDC files. Use proactively after editing any Markdown content.

Tatsh/wiswa-mcp

Agent Claude Code

Fixes mypy type errors and replaces Any with precise types using PEP 695 generics, TypedDicts, Protocols, and Unpack. Use after type-related QA failures.

Tatsh/wiswa-mcp

Agent Claude Code

Expert-level Python author for non-trivial implementation work in the wiswa-mcp project. Use when the task needs idiomatic, production-quality Python beyond a simple edit.

Tatsh/wiswa-mcp

Agent Claude Code

Upgrades older Python patterns to modern 3.12+ syntax (match, walrus, PEP 695 generics, pathlib) while preserving clarity. Use during code cleanup passes.

Tatsh/wiswa-mcp

Agent Claude Code

Runs yarn format followed by yarn qa in a loop, applying fixes until both exit with code 0. Use proactively before every commit and after batches of edits.

Tatsh/wiswa-mcp

Agent Claude Code

Re-runs the Wiswa generator, applies post-processing, verifies the result is safe, and commits. Use when .wiswa.jsonnet or managed templates change.

Tatsh/wiswa-mcp

Agent Claude Code

Generates pytest suites following project conventions (conftest fixtures, parametrize, no docstrings on tests). Use when new code needs tests written from scratch.

Tatsh/wiswa-mcp

Agent Claude Code

Reflects manual edits to Wiswa-managed files back into .wiswa.jsonnet so a regen reproduces them. Use after direct edits to managed files like pyproject.toml.

Tatsh/wiswa-mcp

Agent Claude Code

Extracts embedded Bash from GitHub Actions YAML, runs ShellCheck, and applies fixes. Use after editing workflow scripts.

Tatsh/wiswa-mcp

Settings file Claude Code

Agent settings declaring 63 allowed tools.

Tatsh/wiswa-mcp

Skill Claude CodeCodex

Create high-quality Git commits for all pending changes.

Tatsh/wiswa-mcp

Skill Claude CodeCodex

Prepare and publish a new release for the wiswa-mcp project.

Tatsh/wiswa-mcp

Skill Claude CodeCodex

Run the regen agent from .claude/agents/regen.md.

Tatsh/wiswa-mcp

Instructions file CodexOpenCode

AGENTS.md instructions for Tatsh/wiswa-mcp, covering agents and ai guidance, rules (.claude/rules/), skills (.claude/skills/) and agents (.claude/agents/).

Tatsh/wiswa-mcp

Instructions file

Claude Code instructions for Tatsh/wiswa-mcp: See @README.md for an overview of this project.

Tatsh/wiswa-mcp

MCP server Claude CodeCodexCursor

FastMCP server exposing Wiswa settings discovery for AI assistants. Runs locally from the wiswa-mcp Python package.