honest-test-claims

honest-test-claims is a skill for Claude Code from bhumik154/claim-check. It costs 85 tokens per session (626 once invoked), scanned A, original, MIT.

A rule for reporting test results only from output that was actually seen during the current work session.

In plain words
What is it for?
It helps write accurate summaries, commit messages, and verification reports when discussing tests.
Why use it?
It prevents unsupported claims about test counts or whether a test suite passed.

Skill for Claude Code

Written for Claude Code: shipped in a Claude Code plugin.

Part of the claim-check plugin — 1 skill, 3 hooks shipped together

Good fit It helps write accurate summaries, commit messages, and verification reports when discussing tests.

Compare 6 skills from other repositories ↓
Install with agentmods
npx agentmods add skills/bhumik154/claim-check/honest-test-claims
Install

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.

Any agent
npx skills add bhumik154/claim-check --skill honest-test-claims
Clone the repo
git clone --depth 1 https://github.com/bhumik154/claim-check

Made for: Claude Code.

Or install claim-check, the plugin that ships this one along with the rest of its 1 skill, 3 hooks.

Wrote 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.

agentmods badge for honest-test-claims

README.md
[![agentmods](https://agentmods.dev/badge/skills/bhumik154/claim-check/honest-test-claims/github.svg)](https://agentmods.dev/skills/bhumik154/claim-check/honest-test-claims)
Your own site
<a href="https://agentmods.dev/skills/bhumik154/claim-check/honest-test-claims"><img src="https://agentmods.dev/badge/skills/bhumik154/claim-check/honest-test-claims/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.

agentmods 80×15 button for honest-test-claims

Your own site · 80×15
<a href="https://agentmods.dev/skills/bhumik154/claim-check/honest-test-claims"><img src="https://agentmods.dev/badge/skills/bhumik154/claim-check/honest-test-claims.svg" alt="Reviewed on agentmods" width="80" height="20"></a>
Per session 85 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 626 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe.
Origin original No closer match found in the catalogue.
Token cost

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.

ModelPer sessionOnce invoked
Fable 5.1 $0.00085 $0.00626
Opus 5 $0.00043 $0.00313
Sonnet 5 $0.00017 $0.00125
Haiku 4.5 $0.00009 $0.00063

Measured 11d ago against content hash d738f5107ab1, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-11, from the pricing page.

Security

Grade A, and why

honest-test-claims 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 11d 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.

skills/honest-test-claims/SKILL.md · 64 lines

How it starts

The opening of the file, as written. The whole thing — 64 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Honest test claims

The rule

Never state a test count you did not read from actual output in this session.

If you are about to write "22 passed", "15/17 passing", "all tests pass", or "the suite is green", the number must come from a test run whose output you actually saw. Not from memory of an earlier run, not from the number of tests you believe you wrote, not from a count that was true before your last edit.

If you have not seen the output, you have two honest options:

  1. Run the suite, read the summary line, and state that number.
  2. Say what you actually know: "I added 3 tests; I have not run the suite."

Both are fine. Inventing the number is not.

Why this is a real failure mode, not a hypothetical

Test counts drift constantly during a session. A fix lands, a test splits, a parametrize gains a case, and a number that was true ten minutes ago is now wrong. Restating it from memory feels like reporting and is actually guessing.

The specific trap: a count you stated earlier in the session is evidence to you that the count is right, because you remember asserting it confidently. It is not evidence. It is your own earlier claim.

What counts as having seen the output

The pytest summary line, verbatim, from a run in this session:

=========== 1 failed, 21 passed in 0.09s ===========

That line is authoritative. Do not re-derive a total by counting individual PASSED lines, and do not adjust a previous run's number in your head to account for edits you made since, rerun instead.

Scope matters as much as the number

pytest -k something and pytest tests/unit/ produce real, correct summary lines for a subset. Quoting one of those as "all tests pass" is wrong even though the number is accurate, because the claim is about a scope the run did not cover.

State the scope when it is not the whole suite: "22 passed in tests/unit" rather than "22 passed".

What this skill does not ask

It does not ask you to run the test suite before finishing every task. If you are not making a claim about test results, there is nothing here to satisfy. Say nothing about counts and this skill is irrelevant.

Read the full file on GitHub · 64 lines

Changes

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.

  1. 11d ago First seen · 64 lines · 85 tokens per session scan A d738f5107ab1

Subscribe to this mod's changes

honest-test-claims is a skill published in the GitHub repository bhumik154/claim-check (2 stars, last pushed yesterday), licensed MIT. It adds 85 tokens to every session and 626 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 2026-08-31.