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-initnpx skills add hhamja/claude-code-harness --skill loop-initgit 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 | $0.00022 | $0.00694 |
| Opus 5 | $0.00011 | $0.00347 |
| Sonnet 5 | $0.00004 | $0.00139 |
| Haiku 4.5 | $0.00002 | $0.00069 |
Grade A, and why
loop-init 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 2d ago.
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 — 30 lines — stays where its author put it; the contents beside it link to each section on GitHub.
loop-init — scaffold project-local loop state
Create the mutable loop state under .claude/loop/ in the current project. The plugin itself is immutable; everything the loop learns or tracks lives in these files.
Idempotency: if .claude/loop/ already exists, list its files and stop. Never overwrite existing loop state.
Steps
Scaffold first, questions last: you often cannot tell whether anyone can answer you (headless -p sessions look identical to interactive ones). Creating the files NEVER waits on user input.
-
Detect stack commands for
loop.config.md. Check ecosystem manifests in this order and extract test / lint / build commands:package.json→scripts.test/scripts.lint/scripts.build(prefix with the package manager detected from the lockfile: npm / pnpm / yarn / bun)pyproject.toml→ pytest / ruff / mypy etc. from tool sections or dev dependenciesMakefile→test/lint/buildtargetsCargo.toml→cargo test/cargo clippy/cargo buildgo.mod→go test ./.../go vet ./.../go build ./...
For anything missing or ambiguous, use
TODO: <fill in manually>— do not stop to ask. -
Detect the implementer: run
codex --versionvia Bash. Exit 0 → writeimplementer: codex; anything else →implementer: claude. Never ask — this is a mechanical check. Leavecodex_argsempty either way. -
Create the files from the templates in
references/templates.mdimmediately, using detected values and TODOs. Goal: use the command arguments if given, elseTODO: define the goal; derivebranch:from it (<type>/<slug>, see templates.md). Print one warning line per TODO written. -
Gitignore: ensure the project
.gitignorecontains the line.claude/loop/.*(create.gitignoreif missing; skip if the line already exists). Non-hidden loop files are MEANT to be committed (team sharing, session recovery). -
Interactive refinement (optional): only AFTER the files exist, and only if this session is clearly interactive (a human typed the command and can reply), you may ask ONE question to fill the remaining TODOs, then update the files. If in doubt, skip — TODOs are the designed outcome. Never end the turn with questions instead of created files.
-
Report: created paths + remaining TODOs + the detected implementer + next steps — try
/loopy:loop-run --verify-onlyfirst (grade only), then/loopy:loop-run. Ifimplementer: claudebecause no Codex CLI was found, note: install the Codex CLI and runcodex login, then setimplementer: codexin loop.config.md to enable the cross-model maker/checker split.
What ships with it
1 file beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
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.
- 2d ago First seen · 30 lines · 22 tokens per session scan A 0e2e5df09704
loop-init is a skill published in the GitHub repository hhamja/claude-code-harness (1 stars, last pushed 1mo ago), licensed MIT. It adds 22 tokens to every session and 694 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.