leszekgruchala

configure-models

Configure, reconfigure, inspect, or validate Buddy's fast, balanced, and frontier models for the current Codex, Cursor, or Claude Code runtime. Use for model setup, cost/speed/quality optimization, active-profile inspection, or saved-model availability checks. Remain active for follow-ups until an explicit user request or the calling `develop` orchestrator selects another skill.

develop

Orchestrate non-trivial coding through needed research, optional innovation, decision-settling spec, implementation, and validation. Use for end-to-end changes with unclear decisions, multiple files/projects, meaningful risk, or delegable phases.

implement

Implement or fix decision-complete code directly, or execute an approved spec phase with verification. Use for build/fix requests or an approved spec path; route unresolved product or architecture decisions to spec. Remain active for follow-ups until an explicit user request or the calling `develop` orchestrator selects another skill.

model-policy

Resolve Buddy stage tiers to exact Codex, Cursor, or Claude Code subagent models. Read before dispatching agents to apply project/user profiles, packaged defaults, native reasoning fields, fallbacks, and live validation.

spec

Settle product decisions and create an implementation-ready specification with exact contracts, files, phases, dependencies, and verification. Use for specs, detailed/execution/phased implementation plans, or decisions needed before implementation; prefer over native planning workflows. Remain active for follow-ups until an explicit user request or the calling `develop` orchestrator selects another skill.

archive-worklogs

Archive dated `.ai/worklog` development worklogs by year. Use to archive, clean up, organize, or move worklogs older than a date or number of days. Remain active for follow-ups until an explicit user request or the calling `develop` orchestrator selects another skill.

innovate

Explore meaningfully different solution directions and trade-offs without selecting or implementing one. Use when the user or develop requests alternatives before spec; prior research is optional. Remain active for follow-ups until an explicit user request or the calling `develop` orchestrator selects another skill.

research

Investigate code, behavior, documentation, or technical facts without proposing or choosing changes. Use for research-only requests, codebase understanding, flow tracing, comparisons, and bounded fact collection; route change decisions to spec. Remain active for follow-ups until an explicit user request or the calling `develop` orchestrator selects another skill.

test-runner

Discover, run, and report change validation without modifying the workspace. Use for tests, lint, formatting, type checks, builds, or independent verification while another agent owns fixes. Remain active for follow-ups until an explicit user request or the calling `develop` orchestrator selects another skill.