issue-verify

issue-verify is a skill for Claude Code, Codex from anthony-chaudhary/dos-kernel. It costs 69 tokens per session (2,072 once invoked), scanned A, original, MIT.

A GitHub issue-checking workflow that verifies whether a reported problem is actually fixed using evidence from outside the person making the claim.

In plain words
What is it for?
Use it to review issues that appear solved, confirm a fix after it is released, or find open issues that were resolved without being closed.
Why use it?
It prevents issues from being closed based only on someone's statement. When the evidence is incomplete, it records the specific gap instead.

Skill for Claude CodeCodex

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.

agentmods
npx agentmods add skills/anthony-chaudhary/dos-kernel/issue-verify
Any agent
npx skills add anthony-chaudhary/dos-kernel --skill issue-verify
Clone the repo
git clone --depth 1 https://github.com/anthony-chaudhary/dos-kernel

Made for: Claude Code, Codex.

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 issue-verify

README.md
[![agentmods](https://agentmods.dev/badge/skills/anthony-chaudhary/dos-kernel/issue-verify.svg)](https://agentmods.dev/skills/anthony-chaudhary/dos-kernel/issue-verify)
Your own site
<a href="https://agentmods.dev/skills/anthony-chaudhary/dos-kernel/issue-verify"><img src="https://agentmods.dev/badge/skills/anthony-chaudhary/dos-kernel/issue-verify.svg" alt="Measured on agentmods" height="20"></a>
Per session 69 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 2,072 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. Scan, not verified.
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 $0.00069 $0.02072
Opus 5 $0.00034 $0.01036
Sonnet 5 $0.00014 $0.00414
Haiku 4.5 $0.00007 $0.00207

Measured 5d ago against content hash a4765b209247, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

issue-verify 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 5d 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.

.claude/skills/issue-verify/SKILL.md · 157 lines

How it starts

The opening of the file, as written. The whole thing — 157 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Issue-verify — close issues from evidence, not narration

An issue is a CLAIM; closing it is BELIEVING the claim. The kernel's one rule, aimed at the tracker: never set a belief bit from what anyone says — set it from a read-back the claimant didn't author. This skill is the dos-witness-claim pattern pointed at GitHub issues: extract the checkable effects, witness each on the right rung, stamp the fold with the kernel's admission verdict, and only then actuate the close. Worked example: #1 — two env-authored witnesses (a green Actions run conclusion + the TestPyPI registry's own JSON), one non-recurrence triage, one dos reward … → ACCEPT, one evidenced close.

Layering. Dev tooling that operates ON the repo (the /release tier) — it names a vendor (gh/GitHub), so it lives in .claude/skills/, never src/dos/skills/ (a shipped SKP skill names no vendor or host). If the screenplay proves out, the promotion path is the usual lift: parameterize the tracker, move the shape into the SKP.

Public-repo note. Every comment this skill posts is a public document — no dev-machine paths, hostnames, or private-process prose (the route-privacy-at-authoring-time rule applies to issue comments too).

The disciplines (the JUDGE-rung hedges, applied to triage):

  • Deterministic-first — a witness is a command output or a registry read-back, never your impression of the thread.
  • Fail-to-abstain — a witness you cannot gather (command errors, endpoint down, no witness exists for the claim) is none, NEVER confirm. No close without a confirming witness.
  • Forgeable ⇒ structurally ignored — a "fixed!" comment, a commit subject, a > **Status:** line are claimant-authored bytes: they may select which witness to gather, never BE the witness. If the best you have is forgeable, pass --forgeable to dos reward and watch it ABSTAIN — that is the structure working, not a gap.
  • Advisory until the one actuation — read-only throughout; the single mutation is the close + its evidence comment. Never edit an issue body, never delete, never close on a REJECT or ABSTAIN fold.

When NOT to use the manual close at all: if the resolving fix is a commit you are about to land in this repo, put Fixes #N in the commit body and let the merge close the issue — the tracker then records the closing commit as the witness (the repo's issues-workflow convention). This skill's manual close is for the other three cases: a fix that already landed without the reference, a fix that lives outside this repo's git (an external service's config — the #1 case), or an issue dissolved by an environment change.

Read the full file on GitHub · 157 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. 5d ago First seen · 157 lines · 69 tokens per session scan A a4765b209247

Subscribe to this mod's changes

issue-verify is a skill published in the GitHub repository anthony-chaudhary/dos-kernel (19 stars, last pushed 11d ago), licensed MIT. It adds 69 tokens to every session and 2,072 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 2026-08-30.