Skip to content

execution.md

Latest commit

 

History

History
56 lines (38 loc) · 4.34 KB

execution.md

File metadata and controls

56 lines (38 loc) · 4.34 KB
Raw

Execution — Composition

中文

This document explains the Execution composition: entry, routing, and handoff. Internal step-by-step workflows live in each Skill's SKILL.md and references/.

Skills in this group

All are first-party and self-contained; no install of mattpocock/skills or sol-advisor is required.

Entry → Handoff → Stop

Situation Entry Typical path Handoff / Stop
One clear ticket / SPEC slice implement — user-invoked implement → inspect context → optional agent-config offer (when profile routing, review isolation, or topology materially helps) → execute → verify → hand to review-loop with the right reviewer bounded diff + focused tests + verification evidence; stop at ticket scope
Need to configure model, effort, or execution topology agent-config — model-invoked requires: bounded task or ticket graph + acceptance authority + current host evidence + confirmed profile; determines provider mode and task shape, right-sizes model tier & effort across four execution modes (Cases A, B, C, D) profile-driven execution plan, not execution; executor performs work per plan
Code feature/fix should be test-first tdd — model-invoked red → green → refactor with real tests tests + implementation slice
Hard bug / regression diagnosing-bugs — model-invoked build a tight pass/fail signal → reproduce → hypothesize → instrument → fix → cleanup fix with feedback loop evidence
Merge/rebase conflict resolving-merge-conflicts — model-invoked resolve conflicted files per git guidance clean working tree ready for verification

Two-Repository Architecture: Skill and Companion

The execution configuration system cleanly separates policy reasoning from host persistence and mutation:

  • Skill (skills/agent-config): Installed from this repository (LightDevCoder/skills). Performs task difficulty assessment, tier selection, and execution topology planning. Runs locally in the agent conversation.
  • Companion MCP Runtime: Maintained in the independent repository LightDevCoder/agent-config. Provides 10 native host adapters (Codex, Claude Code, Antigravity / agy, DeepSeek Harness / DSH, OpenCode, ZCode, Cursor, Grok Build, Hermes, Pi) plus generic fallback.
  • Optional Companion: Without the companion MCP server registered, agent-config remains fully operational in session-local, plan-only mode without mutating host configuration. With the companion installed and registered, it enables authentic host capability inspection, profile persistence, configuration preview before apply, and health validation.

Composition with review

implement does not copy reviewer instructions. Coding work follows:

implement → tdd (when appropriate) → code changes + tests → review-loop → code-review

Non-coding work:

implement → artifact → review-loop → generic-review / domain reviewer

The reviewer is read-only; review-loop is the convergence engine; project-review (see review-system) owns final PASS/FAIL/BLOCKED when project acceptance is needed.

When not to use this group