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/0xdarkmatter/claude-mods/loop-opsnpx skills add 0xDarkMatter/claude-mods --skill loop-opsgit clone --depth 1 https://github.com/0xDarkMatter/claude-modsWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/0xdarkmatter/claude-mods/loop-ops)<a href="https://agentmods.dev/skills/0xdarkmatter/claude-mods/loop-ops"><img src="https://agentmods.dev/badge/skills/0xdarkmatter/claude-mods/loop-ops.svg" alt="Measured on agentmods" height="20"></a>What 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 | $0.00086 | $0.05574 |
| Opus 5 | $0.00043 | $0.02787 |
| Sonnet 5 | $0.00017 | $0.01115 |
| Haiku 4.5 | $0.00009 | $0.00557 |
Grade C, and why
loop-ops scanned grade C with 2 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 4d 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.
Downloads and executes remote codehighSupply chain
curl | sh runs whatever the server returns today, which is not necessarily what it returned when this was reviewed.
pre-session files, editing `.claude/`/settings (self-modification), `curl | bash`. Makes network callslowCapability
Not a fault in itself. Listed so you know the mod talks to something, and to what.
pre-session files, editing `.claude/`/settings (self-modification), `curl | bash`. How it starts
The opening of the file, as written. The whole thing — 344 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Loop Ops — outer-loop design discipline
A loop is not a prompt. Turn-by-turn prompting puts you in the loop forever. Loop engineering inverts it: you design a recurring process with memory, verification, and boundaries that discovers work, hands it to agents, verifies the result, and decides — on a schedule or until a goal is met — whether to land it or escalate to a human.
"You shouldn't be prompting coding agents anymore. You should be designing the loops that prompt your agents." — Peter Steinberger
This skill is the outer loop: the orchestration layer above a single agent run. It
is the twin of iterate — iterate is the inner loop (one
metric, one session, git-as-memory); loop-ops is the design discipline for the loop
that schedules and gates inner runs. It does not reimplement spawning or landing; it
composes what this repo already ships.
The six primitives → what owns each here
Every durable loop rests on six primitives. The discipline is wiring them; the parts already exist:
| Primitive | What it is | Owned in claude-mods by |
|---|---|---|
| Schedule | fire the loop on a cadence | native-first: /loop (in-session), Desktop scheduled task (local, unattended), /schedule cloud routines (no local files); /goal is the native completion gate. External (cron/Task Scheduler + loop-run.sh) only for non-Claude-Code control |
| Worktree | isolated, discardable execution context | git-ops worktrees, fleet-worker (per-task worktree) |
| Skills | persistent project knowledge the run loads | this repo's skill layer + your CLAUDE.md |
| Sub-agents | maker/checker separation | Agent/Task; dispatching skills (review, testgen) |
| Connectors | reach tickets / CI / chat | MCP tools, gh, github-ops |
| + State | a durable spine outside the conversation | STATE.md + run-log + budget (this skill) |
The inner improvement loop is iterate; cheap parallel makers are fleet-worker; the
test-gated merge queue is fleet-ops; inter-loop signalling is pigeon. loop-ops is
the doctrine that connects them.
What ships with it
22 files 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.
- assets/examples/pr-watch/github-actions.yml 2.7 KB
- assets/examples/pr-watch/loop-run.sh 1.6 KB runs code
- assets/examples/pr-watch/loop.config.yaml 904 B
- assets/examples/pr-watch/run-log.md 436 B
- assets/examples/pr-watch/run.md 1.7 KB
- assets/examples/pr-watch/STATE.md 963 B
- assets/loop.config.template.yaml 3.4 KB
- assets/model-pricing.json 2.1 KB
- assets/run.sh.template 1.4 KB
- assets/run.template.md 2.2 KB
- assets/STATE.template.md 880 B
- references/claude-code-loops.md 13 KB
- references/failure-modes.md 7.9 KB
- references/pattern-catalog.md 8.0 KB
- references/risk-tiers.md 9.4 KB
- references/state-spine.md 6.4 KB
- scripts/check-pricing-sync.py 7.1 KB runs code
- scripts/loop-check.sh 12 KB runs code
- scripts/loop-doctor.sh 11 KB runs code
- scripts/loop-estimate.py 15 KB runs code
- scripts/loop-scaffold.sh 15 KB runs code
- tests/run.sh 19 KB runs code
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.
- 4d ago First seen · 344 lines · 86 tokens per session scan C 99bbe64ae5ec
loop-ops is a skill published in the GitHub repository 0xDarkMatter/claude-mods (32 stars, last pushed 11d ago), licensed MIT. It adds 86 tokens to every session and 5,574 once invoked, about $0.0004 per session on Opus 5. A static security scan graded it C with 2 findings (downloads and executes remote code, makes network calls). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other skills, from other repositories
x402
Set up Browser Use Cloud payments with x402 — pay per request from a crypto wallet (USDC on Base mainnet), no signup or API key. Two setups it works out up front — "just use it" (set up a wallet so you or Claude Code can run cloud browser tasks paid from the wallet — Claude writes and runs throwaway scripts, nothing…
experiment-plan
Turn a refined research proposal or method idea into a detailed, claim-driven experiment roadmap. Use after research-refine, or when the user asks for a detailed experiment plan, ablation matrix, evaluation protocol, run order, compute budget, or paper-ready validation that supports the core problem, novelty…
omh-context-budget-review
This is a Hermes-native context-budget-review workflow skill.
pinme-uniwebpay
Use when generating, modifying, or reviewing PinMe Worker (Cloudflare Worker TypeScript) code that accepts payments through UniwebPay — payment links, products/prices, checkout sessions, payment status reads, refunds, subscriptions, or handling UniwebPay webhooks with @uniwebpay/sdk in a PinMe project.
presentation-deck
Structure a design presentation for a specific audience and decision. Use when presenting internally. For a portfolio narrative use case-study; for the written argument use design-rationale.
timetracker
Reconstruct daily or weekly project work from dated plans, decisions, meetings, site reports, tasks, and other artifacts, then append user-confirmed manual entries to TIMELOG.md. Use for "log my time," "what did I work on," daily timesheets, or weekly time reconstruction. Never infer hours, start/stop times, billing…