Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx agentmods add skills/hhamja/claude-code-harness/loop-worktreenpx skills add hhamja/claude-code-harness --skill loop-worktreegit clone --depth 1 https://github.com/hhamja/claude-code-harnessWhat it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5 | $0.00020 | $0.00786 |
| Opus 5 | $0.00010 | $0.00393 |
| Sonnet 5 | $0.00004 | $0.00157 |
| Haiku 4.5 | $0.00002 | $0.00079 |
Grade A, and why
loop-worktree scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured yesterday.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 38 lines — stays where its author put it; the contents beside it link to each section on GitHub.
loop-worktree — isolated parallel workers (spawn / integrate)
Implements the loop-engineering skill's references/worktree-guide.md: one work unit = one worktree = one branch. Parallelize ONLY file-disjoint rubric subsets — tasks touching the same files are NOT independent; run those serially in this tree. N worktrees ≈ N× token spend: justify with wall-clock need, not tidiness.
Arguments ($ARGUMENTS): <task> [R# ...] = spawn a worker. --integrate <task> [<task> ...] = fold finished workers back.
Spawn
Requires .claude/loop/ (else: run /loopy:loop-init first, stop).
- Verify every requested
R#exists inrubric.md, and that this subset's implementation files are disjoint from every other planned/live task. Overlap → do not spawn; say why. git worktree add ../<repo>-<task> -b <type>/<task>— cut from the current work branch, never a protected one.- Scaffold the worker's
.claude/loop/in the new worktree:rubric.md: ONLY the requestedR#lines,verify:commands byte-identical.goal.md: one line describing the subset.loop.config.md: copy of this tree's, then setworker: <task>,branch: <type>/<task>,auto_push: false,auto_pr: false— workers never push or open PRs; the orchestrator integrates locally.state.md,memory.md,review.md: fresh from loop-init templates.
- Report the worktree path, branch, and next step: open a NEW session in that directory and run
/loopy:loop-run. Each worktree has its own index/HEAD, soloop_lockandbranch_guardare satisfied independently.
Worker contract (auto_commit's worker mode enforces the commit side): a worker commits its source changes plus .claude/loop/results/<task>.md ONLY — nothing else under .claude/loop/ (its local state.md, subset rubric/goal, and worker config stay uncommitted, so integrating the branch cannot overwrite the orchestrator's loop files). results/<task>.md records: subset criteria pass/fail, files touched, unresolved items.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- yesterday First seen · 38 lines · 20 tokens per session scan A a448203ac2ef
loop-worktree is a skill published in the GitHub repository hhamja/claude-code-harness (1 stars, last pushed 1mo ago), licensed MIT. It adds 20 tokens to every session and 786 once invoked, about $0.0001 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other skills, from other repositories
git-commit-convention
写 git 提交信息、整理提交历史时使用——Conventional Commits 的格式、类型选择和拆分原则。.
delegate-to-deepseek
默认把中等及以下、批量、重复或机械任务作为完整逻辑单元派给 DeepSeek，并由主 Agent 独立验收。适用于批量改文件、扫日志、翻译、ETL、脚本、测试、文档、CRUD、单领域重构、单组件或单 endpoint。主 Agent 可基于上下文和失败代价调整派工策略；用户显式指令、安全、权限、隐私边界和派工后验证不可突破。DEEPSEEKMODE=off 时跳过。.
code-review-checklist
审查代码改动（review diff / PR）时使用——按固定清单过正确性、边界、错误处理和测试盲区，避免只看顺眼不顺眼。.
sruja-project
Procedural workflows for working with the Sruja codebase. Teaches AI editors how to add components, validate changes, and follow patterns.
grill-me
This skill ensures AI and human share a design concept before writing code.
sruja-harness
Verification harness for any AI agent skill. Use with any other skills; before marking a task done, run verify-task. On failure, record a correction learning. Ensures lint, tests, drift, and intent checks pass regardless of which skill generated the code.