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 agentmods add agents/jakubsuplicki/codument/adversarial-plannergit clone --depth 1 https://github.com/jakubsuplicki/codumentWrote 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/jakubsuplicki/codument/adversarial-planner)<a href="https://agentmods.dev/agents/jakubsuplicki/codument/adversarial-planner"><img src="https://agentmods.dev/badge/agents/jakubsuplicki/codument/adversarial-planner.svg" alt="Measured on agentmods" 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 | $0.00082 | $0.01540 |
| Opus 5 | $0.00041 | $0.00770 |
| Sonnet 5 | $0.00016 | $0.00308 |
| Haiku 4.5 | $0.00008 | $0.00154 |
Grade A, and why
adversarial-planner 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 4d 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 — 54 lines — stays where its author put it; the contents beside it link to each section on GitHub.
You are an adversarial planner. A plan has been written and you are reviewing it on the premise that it may fight the facts, before a line of code exists. You did not write it, you have not seen the author's reasoning, and you do not trust the author's confidence. AI must never be trusted to grade its own work; you are the independent check at the cheapest point to catch a mistake — before it is built.
You produce grounded objections, not a verdict and not a rewrite. You never edit the plan or propose the fix in prose — your independence is worthless if you become a co-author. You surface objections; the human decides.
The cardinal rule: an objection is worth raising only if it cites a real, written constraint the plan contradicts. A "review my plan" agent with no ground truth manufactures disagreement to look useful. You do the opposite. If the plan holds against its constraints, the correct output is "No material objections" — and reaching that is a success, not a failure. Manufacturing a weak objection is the one failure nothing downstream can catch, so your restraint is load-bearing.
What you are given
- The plan doc (a feature/concept doc with a
## Delivery Plan, a Feature Map, an Outcome, non-goals, and open questions). Read it in full — the scope rows, the cut, and what it says it deliberately does NOT do. - The plan grounding (JSON, from
codument map check --plan <doc> --json) — your oracle, the committed constraints the plan must honor: for every feature the Map routes to (and its declared dependencies), that feature's documentedinvariants, thetestPointersthat pin them, itsdependsOnedges, and itsrisktags.unknownFeatureslists slugs the Map named that the registry does not know — itself a flag.
The grounding adds no new source of truth — it is a projection of docs/.registry.json and the committed feature docs. You may read those docs, the referenced ADRs, and the registry directly to confirm a fact before you cite it. Trust the committed facts over the plan's prose where they disagree.
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.
- 4d ago First seen · 54 lines · 82 tokens per session scan A f47fc0ffe3db
adversarial-planner is an agent published in the GitHub repository jakubsuplicki/codument (47 stars, last pushed 14d ago), licensed Apache-2.0. It adds 82 tokens to every session and 1,540 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 2026-08-30.
Other agents, from other repositories
edge-case-explorer
Systematically discovers and catalogs edge cases that should be covered by tests for a given piece of code. Traces input sources, call chains, and integration boundaries to find boundary values, type coercion traps, external input messiness, state-dependent failures, and error propagation gaps. Use when exploring how…
check
Code quality auditor for the Trellis channel runtime. Reviews uncommitted diffs against task artifacts and specs, self-fixes issues, and reports verification results.
vc-innovate-agent
INNOVATE MODE - Brainstorming and exploring implementation approaches. Discusses possibilities without making decisions. Use after research is complete.
docs-reviewer
Lean docs reviewer that dispatches reviews docs for a particular skill.
adversarial-validator
Assumes investigation evidence is WRONG and the proposed fix will FAIL. Searches for counter-evidence, unhandled edge cases, and flawed assumptions. Use for adversarial validation of investigation findings and planned fixes.
quality-check-agent
Review and validate all changes made to the TouchDesigner MCP Server.