القائمة
T-shirt sizing for tasks before dispatch. Runs heuristics and optionally LLM classification to determine task size (S/M/L/XL), recommend local vs cloud routing, and set tool budgets.
التثبيت باستخدام Codex أو Claude انسخ هذا Prompt والصقه في Codex أو Claude أو مساعد آخر ليراجع صفحة Skill ويثبّتها لك.
استنادا إلى تصنيف SOC المهني
Branch, commit, push, and open PRs with humanized messages, then monitor GitHub Actions CI. Per-project autonomy gating; blocks direct pushes to the default branch and pushes with a dirty tree. Never auto-merges.
Dispatch patterns for Claude Code as the coding engine. Default harness. Implementation dispatches go through dispatch_coder.py (writes the dispatch receipt the commit gate requires).
Testing strategy, TDD enforcement, spec compliance, and regression checks.
Write implementation plans: bite-sized tasks for coding engine dispatch.
Automated fix loop for failed Quality/Reviewer checks. Parses failures, builds escalating prompts, dispatches through the active harness up to 3 times.
Run independent, file-disjoint plan tasks concurrently, each isolated in its own git worktree + branch. Collects per-task results; never auto-merges.
| name | complexity-triage |
| description | T-shirt sizing for tasks before dispatch. Runs heuristics and optionally LLM classification to determine task size (S/M/L/XL), recommend local vs cloud routing, and set tool budgets. |
| version | 1.0.0 |
| platforms | ["linux","macos","windows"] |
| metadata | {"hermes":{"tags":["triage","complexity","routing","dispatch","planning","t-shirt-sizing"],"related_skills":["local-model-router","writing-plans","architect","skill-discovery"]}} |
Run before planning to classify task complexity and set dispatch parameters.
Before every coding engine dispatch. The triage result determines:
terminal(
command="python3 ~/.hermes-coder/scripts/dynamic_curator.py --task '<task summary>' --repo '<project-dir>'",
workdir="~/.hermes-coder",
timeout=30
)
The script outputs JSON:
size: S, M, L, or XLconfidence: high, medium, or lowmethod: heuristic, llm, or fallbackrouting.engine: "local" or "cloud"routing.model: model identifiertool_budget: max_skills, max_mcp_tools, max_turnsskill_matches: top matched skills to inject| Size | Characteristics | Action |
|---|---|---|
| S | Single file, trivial verb, CSS-only, typo | Skip planning, raw dispatch, local model OK |
| M | Multi-file logic, new component, helper utility | Plan, selective skill injection, cloud model |
| L | New feature, major refactor, multi-file integration | Full plan, full skill injection, cloud model |
| XL | Architecture change, system redesign | Full plan, architect review, cloud model |
Every size dispatches to the coding engine. Size never licenses the coordinator to edit repo files itself — an S task is a small dispatch, not an inline edit. (2026-07-10: inline implementation by the coordinator burned two full turn budgets and shipped below the quality bar.)
The triage LLM pass runs through the active coding harness (resolved from --engine/coding.default_engine). If the harness is unavailable, heuristic-only triage still works (exit code 3). If heuristics are inconclusive and no LLM is available, the script defaults to M sizing with cloud routing.