wyattowalsh/agents

wyattowalsh/agents

Agent

Record agent definition changes with validation evidence for maintainer audit trails.

wyattowalsh/agents

Agent

Run structural eval gates for skills and agents; report adequacy without live LLM runs.

wyattowalsh/agents

Agent

Scaffold for simulating agent permission decisions against sample tool calls (planned).

wyattowalsh/agents

Agent

Scaffold for publishing agent catalog registry artifacts (planned).

wyattowalsh/agents

Agent

Scaffold for transpiling portable agent frontmatter across harness projections (planned).

wyattowalsh/agents

Agent

Verify instruction bridge and discovery parity across harness sync projections.

wyattowalsh/agents

Agent

Review changes for correctness, risk, and maintainability without editing code.

wyattowalsh/agents

Agent

Update or create technical documentation grounded in the current codebase.

wyattowalsh/agents

Agent

Map MCP server tools to harness registry entries and maintainer docs surfaces.

wyattowalsh/agents

Agent

Maintain FastMCP v3 MCP scaffolds in mcp/; align templates with repo conventions.

wyattowalsh/agents

Agent

Coordinate multi-step work by decomposing, delegating, and synthesizing results.

wyattowalsh/agents

Agent

Investigate performance bottlenecks and recommend the highest-leverage fixes.

wyattowalsh/agents

Agent

Audit agent permissionMode, tool allowlists, and OpenCode permission overlays for least privilege.

wyattowalsh/agents

Agent

Create a codebase-grounded implementation plan before coding.

wyattowalsh/agents

Agent

Scaffold for prompt/token optimization reviews on agent and skill bodies (planned).

wyattowalsh/agents

Agent

Prepare release notes, versioning, and ship-readiness checks with cautious permissions.

wyattowalsh/agents

Agent

Investigate a technical question deeply and return a concise evidence-backed summary.

wyattowalsh/agents

Agent

Audit code and configuration for security risks without making changes.

wyattowalsh/agents

Agent

Read-only advisor for skill authoring, audits, and eval planning. Does not edit skill files.

wyattowalsh/agents

Agent

Classify incoming work by severity, harness surface, and ownership; route to specialist agents.

wyattowalsh/agents

Plugin Claude Code

Portable AI agent skills, subagents, MCP configuration, and shared workflow guidance.

wyattowalsh/agents

Hook Claude Code

Runs after a tool call finishes for Edit, Write, WebSearch, web_search, fetch, fetcher, trafilatura, docling, mcp__.* and mcp_.* tool calls, running an inline shell check (2 commands). From wyattowalsh/agents.

wyattowalsh/agents

Hook Claude Code

Runs before the agent uses a tool for Bash, bash, run_shell_command, Shell, Edit, Write, MultiEdit, edit, create, write_file, replace, Bash, bash, run_shell_command, shell, terminal, analyze_image, attach_image, screenshot, upload_image, view_image, Read, read, Write, Edit, MultiEdit, apply_patch, edit, create…

wyattowalsh/agents

Hook Claude Code

Runs when you submit a prompt, before the agent sees it, running ./hooks/run-wagents-hook research-prompt-triage-context --ha. From wyattowalsh/agents.