Borrowing it
Nothing to install: this file belongs to moberghr/mtk-agent-toolkit. 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/moberghr/mtk-agent-toolkit/main/.claude/skills/verification-before-completion/SKILL.mdgit clone --depth 1 https://github.com/moberghr/mtk-agent-toolkitWrote 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/moberghr/mtk-agent-toolkit/verification-before-completion)<a href="https://agentmods.dev/skills/moberghr/mtk-agent-toolkit/verification-before-completion"><img src="https://agentmods.dev/badge/skills/moberghr/mtk-agent-toolkit/verification-before-completion/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/moberghr/mtk-agent-toolkit/verification-before-completion"><img src="https://agentmods.dev/badge/skills/moberghr/mtk-agent-toolkit/verification-before-completion.svg" alt="Reviewed on agentmods" width="80" 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 5.1 | $0.00027 | $0.04446 |
| Opus 5.5 | $0.00011 | $0.01778 |
| Sonnet 5.5 | $0.00005 | $0.00889 |
| Haiku 4.5 | $0.00003 | $0.00445 |
Grade B, and why
verification-before-completion scanned grade B 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 16d 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.
Reads agent configuration directoriesmediumAgent snooping
.claude/, .codex/, .gemini/ hold keys, settings and other credentials a mod has no legitimate need for.
cat .claude/tech-stack 2>/dev/null || echo "(not set)" Makes network callslowCapability
Not a fault in itself. Listed so you know the mod talks to something, and to what.
| SC2 | verified | `curl :5080/health → 200, body {"status":"ok"}` | How it starts
The opening of the file, as written. The whole thing — 241 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Verification Before Completion
Active Stack
echo "--- Tech Stack ---"
cat .claude/tech-stack 2>/dev/null || echo "(not set)"
Overview
No completion claim is valid without fresh evidence from an actual command execution. "Should work," "probably fixed," and "looks correct" are not verification. Run the command, read the output, check the exit code, then state the result with evidence.
Verification contract. Every success criterion carries an evidence_channel (the surface where the observable result is captured) and an observable (a binary pass/fail statement declared before execution). Verification must go criterion-by-criterion — trust the list over memory. For behavior-shaped changes, tests alone never prove done: the evidence_channel must be a real execution surface (smoke-boot, http-probe, db-state-diff, cli-stdout, or browser), not a build artifact.
Re-arm rule. Any edit that lands after the most recent verification resets every criterion's status to re-armed. A completion claim is rejected while any criterion is re-armed. Re-verification must run after the edit, cite the observable result per criterion, and set each criterion to verified before the claim is accepted.
Frozen Criteria & Tamper Check
The success_criteria[] definitions (id, observable, evidence_channel) are frozen at Phase 2.5 approval. They are the goalposts; the run is not allowed to move them. The criteria are read-only ground truth during implementation — the same invariant F15 (Frozen-Replay / Non-Varying Evidence) names: editing what measures you to move the number defeats the gate.
Before accepting any completion claim, run the tamper check: confirm the frozen criteria have not changed since approval. Spec sidecars are gitignored working artifacts (spec-driven-development ignores docs/specs/* except baseline/), so there is no committed version to diff. Instead, the implement Phase 2.5 gate records an approval seal on the workflow artifact: a SHA-256 over the approved spec and plan bodies, plus a hash of the sidecar's frozen success_criteria fields (id, observable, evidence_channel). verify-seal re-derives both from disk and compares:
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.
- 16d ago Changed · -26 lines 9c98cd242898
- 1mo ago Changed · +12 lines e93a54138536
- 1mo ago Changed · -15 lines 84b3b46519fc
- 2mo ago First seen · 270 lines · 27 tokens per session scan B b38669b156ea
verification-before-completion is a skill published in the GitHub repository moberghr/mtk-agent-toolkit (9 stars, last pushed 8d ago), licensed MIT. It adds 27 tokens to every session and 4,446 once invoked, about $0.0001 per session on Opus 5.5. A static security scan graded it B with 2 findings (reads agent configuration directories, makes network calls). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other skills, from other repositories
research-engineer
An uncompromising Academic Research Engineer. Operates with absolute scientific rigor, objective criticism, and zero flair. Focuses on theoretical correctness, formal verification, and optimal implementation across any required technology.
tika-eval-compare
Compare extracts from two Tika builds over a corpus to detect regressions in content, encoding, exceptions, and embedded-document handling. Use for "compare before/after extracts", "eval this change against the corpus".
cua-driver
Use Cua Driver for desktop or browser tasks that are awkward or unavailable through Bash/APIs, or when the user explicitly wants GUI interaction: app testing, visual bug reproduction, form filling, calendar entry, screenshots, and demo recording. Also covers Cua setup; not OpenAI Codex Computer Use or web research.
jetson-validate-image
Use after jetson-flash-image to run static BSP checks, on-target smoke/regression tests on a flashed DUT, or both. Not for build or flash steps. Triggers: validate bsp, on-target validation.
atmos-validation
Validate Atmos projects, components, arbitrary JSON Schema inputs, EditorConfig, and GitHub Actions; use affected-file selection and native CI annotations.
skill-benchmark
Benchmark AI skill effectiveness by measuring implementation quality against legacy constraints.