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.
git clone --depth 1 https://github.com/yuri-semenenko/ai-engineering-workspaceWrote 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/commands/yuri-semenenko/ai-engineering-workspace/pr-recheck)<a href="https://agentmods.dev/commands/yuri-semenenko/ai-engineering-workspace/pr-recheck"><img src="https://agentmods.dev/badge/commands/yuri-semenenko/ai-engineering-workspace/pr-recheck/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/commands/yuri-semenenko/ai-engineering-workspace/pr-recheck"><img src="https://agentmods.dev/badge/commands/yuri-semenenko/ai-engineering-workspace/pr-recheck.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 | $0.00026 | $0.01192 |
| Opus 5.5 | $0.00010 | $0.00477 |
| Sonnet 5.5 | $0.00005 | $0.00238 |
| Haiku 4.5 | $0.00003 | $0.00119 |
Grade A, and why
pr-recheck 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.
How it starts
The opening of the file, as written. The whole thing — 98 lines — stays where its author put it; the contents beside it link to each section on GitHub.
PR Recheck Prompt
Role: Act as a Staff-level code reviewer doing a follow-up pass on a PR that has already been reviewed once and has since received fixes.
Context: A prior round of inline comments exists. The author has pushed changes. The job is to check whether each open comment is now resolved, not to redo the first review.
Task:
-
Confirm
gh auth statusand that the PR is still open (gh pr view <id> --json state,mergedAt). -
Load the open review threads. Thread resolution state is only available via GraphQL, not REST:
gh api graphqloverreviewThreads, keepingisResolved: false. -
Read the new commits and current diff, then record three separate things for each open thread, with the file:line and the commit that addresses it:
- Severity, from the first review: Critical / Important / Optional. When the thread does not state it, classify it from what it describes and mark the severity as inferred.
- Status:
addressed/partial/not addressed/wont-fix (author replied). Awont-fixreply is the author's position, not the reviewer's acceptance. - Blocks approval:
yes,no, orno (accepted risk). A Critical or Important thread blocks unless it isaddressedor its risk is accepted. An Optional thread does not block.
Accepting a risk is a decision about what remains, recorded with its reason: why the risk is small, where it is mitigated, or the follow-up ticket. On a Critical or Important item, thread or new finding alike, the decision is the user's: propose it with the reason, keep the item blocking, and mark it
no (accepted risk)only after the user says yes. Record the agreed reason as a drafted reply on the thread, posted on confirmation like any comment.addressedmeans the scenario the thread described no longer happens, not that the cited line changed; a regression test added for it counts only if it would fail without the fix. Read it against the fix, or, when that is cheap, run it with the fix taken out in an isolated copy of the PR head (a scratch copy orgit worktree add --detach, never your working tree or the PR branch, nothing committed or pushed). When that cannot be done safely, or the run errors or times out, say the failure was read, not observed. -
Scan the changed code for new findings, classified Critical / Important / Optional, including regressions the fix itself introduced on adjacent paths. A new Critical or Important finding blocks approval unless its risk is accepted the same way; a new Optional one does not.
Constraints:
- Write in English unless explicitly requested otherwise.
- Plain prose in any posted comment. No arrow, tilde, or em-dash symbols. No AI-sounding phrasing.
- Anti-nitpick gate: drop Optional findings whose cost of ignoring is none.
- Cite file and line for every point.
- Gather then judge: read the open threads and the new work first, then build the matrix and rule. The gathering pass is evidence, not the verdict.
Decision gate:
- Clean path: if nothing blocks approval (no thread and no new finding is marked
yesin steps 3 and 4), resolve theaddressedthreads (GraphQLresolveReviewThread; resolution is reversible) and leave every other thread open. Open Optional threads and new Optional findings do not change this path; list them as non-blocking, with any new ones drafted as comments. Do NOT auto-approve: approving a PR is an outward action, so state that the PR is clean and wait for an explicit yes before runninggh pr review <id> --approve. Show the non-blocking drafts with that question: approving and posting them are separate yeses, and an approval alone posts nothing. Print the matrix and the lineall clear - N threads resolved, M non-blocking items open; ready to approve. - Blocked path: otherwise, when anything blocks approval, do not approve and do not resolve unaddressed threads. Output the matrix, the new draft comments, and any proposed risk acceptance with its reason, then wait for confirmation before posting or resolving anything. Resolution and approve are separate confirmations. A confirmation covers the drafts it names; a finding added after it is confirmed on its own. An explicit instruction earlier in the chat to post this recheck's comments is that confirmation; do not ask for it twice.
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.
- 5d ago Changed · +35 lines 03d78e47c270
- 6d ago First seen · 63 lines · 26 tokens per session scan A a4b8ab4b93fb
pr-recheck is a command published in the GitHub repository yuri-semenenko/ai-engineering-workspace (4 stars, last pushed 3d ago), licensed MIT. It adds 26 tokens to every session and 1,192 once invoked, about $0.0001 per session on Opus 5.5. 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 2026-10-31.
pr-recheck compared with commands from other repositories
| Mod | Type | Stars | Last push | Licence | Scan | Tokens | Works with |
|---|---|---|---|---|---|---|---|
| pr-recheckyuri-semenenko/ai-engineering-workspace | Command | 4 | 3d ago | MIT | A | 26 | GitHub Copilot |
| pr-quality-gate-qarjmurillo/ai-agents | Command | 49 | yesterday | MIT | A | 0 | GitHub Copilot |
| pr-quality-gate-agent-safetyrjmurillo/ai-agents | Command | 49 | yesterday | MIT | A | 0 | Claude Code, GitHub Copilot |
| pr-quality-gate-code-qualityrjmurillo/ai-agents | Command | 49 | yesterday | MIT | A | 0 | GitHub Copilot |
| pr-quality-gate-decision-rigorrjmurillo/ai-agents | Command | 49 | yesterday | MIT | A | 0 | GitHub Copilot |
| pr-quality-gate-reliabilityrjmurillo/ai-agents | Command | 49 | yesterday | MIT | A | 0 | GitHub Copilot |
| pr-quality-gate-roadmaprjmurillo/ai-agents | Command | 49 | yesterday | MIT | A | 0 | GitHub Copilot |