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 daronthedragon/skillsmith --skill prove-itgit clone --depth 1 https://github.com/daronthedragon/skillsmithWrote 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/daronthedragon/skillsmith/prove-it)<a href="https://agentmods.dev/skills/daronthedragon/skillsmith/prove-it"><img src="https://agentmods.dev/badge/skills/daronthedragon/skillsmith/prove-it/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/daronthedragon/skillsmith/prove-it"><img src="https://agentmods.dev/badge/skills/daronthedragon/skillsmith/prove-it.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 | $0.00113 | $0.00803 |
| Opus 5.5 | $0.00045 | $0.00321 |
| Sonnet 5.5 | $0.00023 | $0.00161 |
| Haiku 4.5 | $0.00011 | $0.00080 |
Grade A, and why
prove-it scanned grade A with 1 finding 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 2mo 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.
Makes network callslowCapability
Not a fault in itself. Listed so you know the mod talks to something, and to what.
1. Name what would prove the claim: a test command, a build, a curl, a script, How it starts
The opening of the file, as written. The whole thing — 79 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Prove It
An agent's most expensive failure is a confident claim that turns out false: "fixed", "should work now", "the contributor list is clean" - said before anyone looked. The cost is not the bug; it is the user trusting the claim, moving on, and finding out later. This skill makes every claim carry its receipt.
Persistence
ACTIVE EVERY RESPONSE once triggered. Still active when the change feels too small to test. Off only when the user says "stop prove-it" or "normal mode".
Procedure
Run this before every completion message, in order.
- Name what would prove the claim: a test command, a build, a curl, a script, opening the page, reading the file back. If nothing can prove it, go to 5.
- Run it. Do not describe what running it would show - run it.
- Read the real output. If it contradicts the claim, the claim is false; fix the work, not the wording, and return to 1.
- In the completion message, paste the command and the decisive lines of its output next to the claim it supports. One receipt per claim.
- If a claim cannot be run - no key, no network, no access - write the word "unverified" beside it and say what would verify it. Never upgrade "unverified" to "works" by rephrasing.
- Before sending, scan the message for the phrases in Rules. Delete or replace each one with a receipt or with "unverified".
Rules
- The phrases "should work", "looks correct", "this will", "I believe this fixes", and "appears to" do not appear in a completion message. Each is replaced by a command and its output, or by "unverified".
- A claim about a remote state (a web page, a repo, an API, a deployment) is verified by reading that remote state, not the local files that produced it. Local clean and remote clean are different facts.
- One check per claim. Verifying the API does not verify the page; verifying the commit message does not verify the contributors panel.
- When a check fails, the completion message reports the failure and its output. Silence about a failed check is a false claim.
- The receipt is the decisive lines, not a log dump. Trim to what settles it.
What ships with it
4 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.
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.
- 2mo ago First seen · 79 lines · 113 tokens per session scan A f980857cc298
prove-it is a skill published in the GitHub repository daronthedragon/skillsmith (2 stars, last pushed 2mo ago), licensed MIT. It adds 113 tokens to every session and 803 once invoked, about $0.0005 per session on Opus 5.5. A static security scan graded it A with 1 finding (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
systematic-debugging
4-phase root cause debugging: understand bugs before fixing.
module-regression
Govern cross-module regression in large repositories with a REGRESSION.md ledger that records each module's downstream consumers and executable acceptance command. After a change, run the changed module and every affected downstream command, and use observed exit codes as regression evidence under the project's…
fix-failing-tests
Diagnose a failing test in the googleapis/mcp-toolbox repo and land a fix by reasoning from the actual error: read the failure, reproduce it, shrink it until the cause is forced into the open, then fix the cause. Use this whenever a test or CI job is red, a build breaks after a change, many packages fail at once, or a…
reproduce-bug
Reproduce a reported bug in googleapis/mcp-toolbox and decide whether it is real, delivering an evidence-backed verdict: confirmed, already fixed, misconfiguration, client-side, works as intended, not reproducible, or blocked. Use whenever a maintainer asks you to reproduce, verify, confirm, or investigate a bug…
credit-note-fixer
Fix the tiny credit-note formatting bug and rerun the exact targeted test command.
verify-omnigent
Drive Omnigent the way a user does and prove a behavior with recorded evidence, using an isolated server, runner, and mock model plus a feature map of every user entry point. Load before reproducing a user-facing bug, before claiming a fix works, or when reviewing whether a change covered every surface (session vs.…