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 skills add OutlineDriven/odin-claude-plugin --skill verification-before-completiongit clone --depth 1 https://github.com/OutlineDriven/odin-claude-pluginWrote 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/outlinedriven/odin-claude-plugin/verification-before-completion)<a href="https://agentmods.dev/skills/outlinedriven/odin-claude-plugin/verification-before-completion"><img src="https://agentmods.dev/badge/skills/outlinedriven/odin-claude-plugin/verification-before-completion.svg" alt="Measured on agentmods" 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 | $0.00057 | $0.01005 |
| Opus 5 | $0.00028 | $0.00502 |
| Sonnet 5 | $0.00011 | $0.00201 |
| Haiku 4.5 | $0.00006 | $0.00101 |
Grade A, and why
verification-before-completion 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 — 47 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Verification before completion
Contract
| Field | Bound contract |
|---|---|
| Trigger | A done, fixed, passes, complete, or ready claim is imminent; a commit, PR, or next-task move is about to happen; or satisfaction is about to be expressed. |
| Authority | Read-only. No file, VCS, credential, paid, published, deployed, or remote mutation. Reads only what is required to classify the claim, and executes the one named proving action. |
| Side effect | Runs the narrow proving command or scenario; edits nothing; blocks the completion claim when the output contradicts it. |
| Done | Each claim carries fresh output from exactly one proving action (or the UNVERIFIED-NO-SURFACE classification when no checkable surface exists), the claim wording matches the output, and failures are reported honestly. |
Inputs
- Fires on: Explicit invocation, or a user utterance that contains a completion claim.
- Must supply: The specific claim being made; what would prove it; whether it has been run; the output of the check.
- Optional: The exact command or scenario run; the raw output read; any partial results.
Procedure
- Handle the no-check surface. If the change has no checkable surface (comment-only, pure prose, a rename with no attached behavior), state that plainly and classify as UNVERIFIED-NO-SURFACE. Do not invent a check to run; skip the proving-action steps and report the classification. Done when: the surface is classified as checkable or UNVERIFIED-NO-SURFACE.
- Name the proving action. For a checkable surface, identify the precise action that would prove the claim: a specific test, build, lint pass, input scenario, file read, or reproduction of the original failure. "The three tests covering the retry path" is precise; "the test suite" is not. Done when: one precise proving action is named for the checkable surface.
- Run it. Execute the named action after the last relevant edit to the code. Do not run before the edit and claim the result as post-edit evidence. Done when: the proving action has been executed after the last edit.
- Read all of it. Read the full output and the exit code. Do not read only a tail, a summary, or the last line. Treat the exit code as a fact to check. Done when: the full output and exit code are read.
- Classify. VERIFIED if the output confirms the claim at its stated scope; PARTIAL if evidence exists but does not cover the full scope (state what passed and what was left unchecked); UNVERIFIED if no check was run (state "not run" plainly); FAILED if the output contradicts the claim (report the contradiction exactly). Done when: one classification is assigned with its justification.
- Report. State the classification, the action that was run, what it showed, and the exact claim wording it does or does not support. Do not hedge; do not round up a partial result to "done." Done when: the report states the classification, action, output, and claim wording.
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.
- yesterday Changed · -33 tokens per session 71bdde6200dd
- 3d ago First seen · 47 lines · 90 tokens per session scan A ce2f18eb8226
verification-before-completion is a skill published in the GitHub repository OutlineDriven/odin-claude-plugin (35 stars, last pushed today), licensed Apache-2.0. It adds 57 tokens to every session and 1,005 once invoked, about $0.0003 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 2026-09-04.
Other skills, from other repositories
systematic-debugging
Use when encountering any bug, test failure, or unexpected behavior, before proposing fixes.
local-ai-agents
Build local-first AI agents that run entirely on a developer workstation with Microsoft Foundry Local and Qwen function-calling models. Covers Small Language Models (SLMs), the OpenAI-compatible local endpoint, sandboxed local tools, local RAG with Chroma, local MCP servers, hybrid cloud/local routing, and the…
next-cache-components-adoption
Turn on Cache Components in a Next.js app and resolve the blocking routes it surfaces. Use when the user wants to enable, adopt, or migrate to Cache Components, flip the cacheComponents flag, work through a flood of blocking-prerender / instant validation errors, run the cache-components-instant-false codemod, or…
next-partial-prefetching-adoption
Turn on Partial Prefetching in a Next.js app and work through the insights it surfaces. Use when the user wants to enable or adopt Partial Prefetching, flip the partialPrefetching flag, opt routes in with export const prefetch = 'partial', audit Link prefetch={true} behavior, preserve existing prefetched UI with…
chronicle
Analyze Copilot session history for standup reports, usage tips, session search, and session reindexing. Use when the user asks for a standup, daily summary, usage tips, workflow recommendations, wants to search or find past sessions by keyword/file/PR, wants to reindex their session store, or asks about deleting…
babysit-pr
Babysit a GitHub pull request after creation by continuously polling review comments, CI checks/workflow runs, and mergeability state until the PR is merged/closed or user help is required. Diagnose failures, retry likely flaky failures up to 3 times, auto-fix/push branch-related issues when appropriate, and keep…