Borrowing it
Nothing to install: this file belongs to IgorGanapolsky/ThumbGate. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.
curl -O https://raw.githubusercontent.com/IgorGanapolsky/ThumbGate/main/.claude/skills/scope-discipline/SKILL.mdgit clone --depth 1 https://github.com/IgorGanapolsky/ThumbGateWrote 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/igorganapolsky/thumbgate/scope-discipline)<a href="https://agentmods.dev/skills/igorganapolsky/thumbgate/scope-discipline"><img src="https://agentmods.dev/badge/skills/igorganapolsky/thumbgate/scope-discipline/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/igorganapolsky/thumbgate/scope-discipline"><img src="https://agentmods.dev/badge/skills/igorganapolsky/thumbgate/scope-discipline.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector pass
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 5.1 | $0.00082 | $0.00785 |
| Opus 5 | $0.00041 | $0.00392 |
| Sonnet 5 | $0.00016 | $0.00157 |
| Haiku 4.5 | $0.00008 | $0.00078 |
Grade A, and why
scope-discipline 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 10d 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 — 72 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Scope Discipline — Close Loops, Don't Open Them
The failure this prevents
A single request balloons into a chain where every step uncovers new work and you chase it immediately, so nothing reaches "done + verified + deployed." This session: RAG → OAuth → directory submission → reviewer key → consent-key validation → CodeQL fixes — six threads, each started before the prior landed. The result feels like motion but lands nothing. (2026 reliability research flags "repetitive actions / no task completion" as the top long-horizon failure trigger — a metacognitive check on this is recommended: arxiv 2509.19783.)
The rule: one in-flight task, with a parking lot
- One task is
in_progressat a time (mirror it in the task-state-ledger). - When a new issue surfaces mid-task, do NOT start it. Write it to a parking lot and keep going. Park, don't pivot.
- A task is closeable only when verified (the
evidence-first-answer/ deployment-gate bar), not when code is written. - After closing, pick the next item from the parking lot or ask the CEO which to take if priority is ambiguous.
The parking lot
Keep a ## Parking lot section in the task-state-ledger:
## Parking lot (found mid-task — do NOT start until current loop closes)
- [security] CodeQL: 2 critical cmd-injection + 1 XSS on main — own PR
- [infra] THUMBGATE_REVIEWER_KEY not set on Railway
- [docs] /docs/connectors 404, favicon 404
For genuinely independent work, prefer spawning a separate task (e.g. the
spawn_task chip) over context-switching this session.
When the user says "continue" / "do everything"
That means make progress on the current loop, not fan out into new threads. The most valuable thing is the next closed loop, not the most new activity. Before branching, ask: "does this finish the current task or start a new one?" If it starts a new one → park it.
Allowed exception
A discovery that makes the current task unsafe to finish (e.g. you find the thing you're about to ship has a security hole) DOES belong in the current loop — fixing it IS closing the loop correctly. The test: "is this a prerequisite of the current Goal, or a new Goal?" Prerequisite → handle now. New Goal → park.
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.
- 10d ago First seen · 72 lines · 82 tokens per session scan A b4287a5a4712
scope-discipline is a skill published in the GitHub repository IgorGanapolsky/ThumbGate (26 stars, last pushed today), licensed MIT. It adds 82 tokens to every session and 785 once invoked, about $0.0004 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 2026-08-30.
Other skills, from other repositories
agentguard
Runtime guardrails for AI coding agents. Stop loops, budget overruns, retry storms, and timeouts before they burn money. Zero dependencies, local-first, MIT licensed.
crm-pipeline-playbook
How to answer pipeline, renewal, and account-health questions using the CRM tools.
deciding-who-decides
Decides who holds authority for a change — the agent at the edge or a human gate — by matching decision rights to reversibility, evidence, and consequence, and names the escalation trigger. Use when an agent could act on something irreversible, trust-bearing, or thinly evidenced. Do not use for trivial reversible…
staying-on-mission
Tests the current work against a written, lasting mission anchor and forces one choice, re-anchor, escalate, or stop. Use when an agent keeps finishing tasks but the work drifts from the goal, scope creeps, the same action repeats in a loop, or standards slip one small step at a time. Do not use for a tiny edit with…
verifying-final-artifacts
Verifies the delivered artifact a consumer actually receives (a figure, chart, PDF, SVG, screenshot, compiled build, or deployed response) by regenerating it fresh and directly observing the output, not the source that produces it, and ends in one honest verdict. Use when the deliverable is a produced artifact and…
closing-stale-packets
Brings an abandoned or half-filled change packet to an honest terminal state by completing it, closing it with a recorded rationale, or deleting it. Use when ng status flags a scaffold or invalid packet, a long session left a draft behind, or work was abandoned mid-packet. Do not use for an actively in-progress…