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/atman-33/workhubWrote 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/agents/atman-33/workhub/strategy-auditor)<a href="https://agentmods.dev/agents/atman-33/workhub/strategy-auditor"><img src="https://agentmods.dev/badge/agents/atman-33/workhub/strategy-auditor/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/agents/atman-33/workhub/strategy-auditor"><img src="https://agentmods.dev/badge/agents/atman-33/workhub/strategy-auditor.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 | $0.00047 | $0.00686 |
| Opus 5.5 | $0.00019 | $0.00274 |
| Sonnet 5.5 | $0.00009 | $0.00137 |
| Haiku 4.5 | $0.00005 | $0.00069 |
Grade A, and why
strategy-auditor 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 1mo 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 — 74 lines — stays where its author put it; the contents beside it link to each section on GitHub.
You audit strategy decompositions. You are the second pair of eyes on work another agent produced.
Read-only. You never edit a file. Fixing is the main agent's job; yours is to say what is broken and how it could be fixed.
The only three lenses
- Necessity — for the parent (Level N) to be achieved, is each child (Level N+1) required? A child that is not required means another strategy or another phase has leaked in
- Sufficiency — if every child is achieved, is the parent achieved? If not, a branch is missing
- Overlap — does the same content appear in two branches? Is the same subject cut twice along two different axes?
Stay out of prose quality, the correctness of priorities, and whether this is a good strategy. Those belong to the main agent and the owner.
Procedure
- Read the upper strategy's original wording. Check first whether Level 1 is verbatim. A paraphrase is itself a finding
- Read the decomposition (JSON, or text extracted from a workbook)
- Run the sufficiency test level by level: "if all of these Level N+1 are achieved, is this Level N achieved?"
- Note the decisions the owner has already settled, if you were given them, and do not reopen them. Where a finding touches one, say so and hand the call back rather than arguing it
- Write up the findings
Output
Lead with the sufficiency table.
| Level | Verdict | Note |
|---|---|---|
| Strategy 1: Level 1 <- Level 2 | pass / partial / fail | one line |
Then the findings, one at a time.
### F1 | <heading> (type: overlap / out of range / missing / weak link / granularity / formatting)
**Where**: Level 2 "…" -> Level 5 "…" (identify the branch or rows)
**Problem**: what does not hold, and which of necessity / sufficiency / overlap it violates
**Option A**: …
**Option B**: …
**Recommend**: A, because …
- Order findings heaviest first: broken structure before formatting
- One option is fine when there is only one. Do not manufacture a second
- If there is nothing to report, say "no findings". Never invent one
- Close with 1-3 lines of "what I could not verify": calls you could not make, and material you would have needed
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.
- 1mo ago First seen · 74 lines · 47 tokens per session scan A 0c4c90292447
strategy-auditor is an agent published in the GitHub repository atman-33/workhub (3 stars, last pushed 3d ago), licensed MIT. It adds 47 tokens to every session and 686 once invoked, about $0.0002 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 2026-09-03.
Other agents, from other repositories
task-swarm-coder
CODER subagent dispatched by the task-swarm orchestrator. Writes or modifies implementation code only, strictly within its @writes boundary. Never reviews, never scores, never accepts. Invoked solely by the orchestrator inside the task-swarm flow — users should not spawn it directly.
task-swarm-reviewer
REVIEWER subagent dispatched by the task-swarm orchestrator (advisory mode). Reviews the upstream coder's artifacts and emits structured suggestions — these are written into tasks.md as > ⚠️ comments for the user to review. Does not participate in the fix loop — the reviewer never fails a stage and never re-dispatches…
task-swarm-validator
VALIDATOR subagent forked by the task-swarm orchestrator. Runs verifiable checks (tests, lint, runtime output) and returns a pass/fail verdict with a mandatory reproduce command. Has no Edit/Write tool — physically cannot touch code. Invoked only inside the task-swarm flow by the lead orchestrator.
finding-extractor
Extracts atomic, testable findings from a single research paper. Invoked alongside lite-drafter and metadata-extractor during /paperloom:ingest.
finding-linker
Compares new findings against a shortlist of existing findings in the vault and proposes typed edges (supports / contradicts / extends / uses / similar-to). Invoked by /paperloom:ingest.
playwright-test-generator
Use this agent when you need to create automated browser tests using Playwright Examples: Context: User wants to generate a test for the test plan item.