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 skills/tiltely/lens/adversarynpx skills add Tiltely/lens --skill adversarygit clone --depth 1 https://github.com/Tiltely/lensWrote 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/skills/tiltely/lens/adversary)<a href="https://agentmods.dev/skills/tiltely/lens/adversary"><img src="https://agentmods.dev/badge/skills/tiltely/lens/adversary.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 | $0.00098 | $0.00785 |
| Opus 5 | $0.00049 | $0.00392 |
| Sonnet 5 | $0.00020 | $0.00157 |
| Haiku 4.5 | $0.00010 | $0.00078 |
Grade A, and why
adversary 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 3d 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 — 60 lines — stays where its author put it; the contents beside it link to each section on GitHub.
The Lens of the Adversary
Read first: ${CLAUDE_PLUGIN_ROOT}/core/protocol.md and ${CLAUDE_PLUGIN_ROOT}/core/dossier.md.
This is a META lens: it examines the PLAN itself for weakness, not a domain perspective. It runs as the mandatory final step of /lens:socratic (and standalone on any plan/diff). Adversarial critique reliably surfaces what normal planning misses — so a plan is not "done" until it survives this pass.
First: suggest plan mode, then WAIT
If the work is substantial enough to deserve a written plan, SUGGEST to the user that they enter Claude Code's plan mode (or otherwise formalize the plan as a written artifact) — and WAIT for their input. Do NOT toggle plan mode yourself. If they decline, or the work is small, critique the plan as it stands.
The red-team battery
Attack the plan from these angles; every finding must be CONCRETE, not generic:
- Refutation: name the ONE assumption that, if wrong, breaks the plan. Is it verified or merely assumed? An assumed load-bearing assumption is the top risk.
- Weakest decision: of the decisions recorded in the dossier/plan, which is least defensible, and why? What alternative was dismissed too fast?
- What discovery missed: re-run the four excavations (protocol.md) AGAINST the plan — a caveat, rabbit hole, blast-radius entry, or platform-matrix gap the plan does not address. This is where socratic plans most often leak.
- Reversibility & blast radius: what here is hard to undo, and is the plan treating it with the caution it deserves? Who/what breaks if this ships wrong?
- "First thing that worked" test: is this the best plan, or the first coherent one? Name one genuinely different approach and say why it lost — or didn't.
On substantial work, fan out: run the angles as independent adversarial passes (or read-only subagents) so no single line of attack dominates, then merge.
Disposition — every surviving finding
Each finding the plan cannot already answer becomes exactly one of:
- revise — change a decision in the dossier (state the new decision);
- reopen — spawn a frontier branch (protocol.md) and hand back to discovery;
- accept — record as an explicit accepted risk, with an owner, in the dossier. Never let a real finding evaporate as "noted".
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.
- 3d ago First seen · 60 lines · 98 tokens per session scan A 592edebb942c
adversary is a skill published in the GitHub repository Tiltely/lens (5 stars, last pushed 25d ago), licensed MIT. It adds 98 tokens to every session and 785 once invoked, about $0.0005 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.
Other skills, from other repositories
systematic-debugging
Use when encountering any bug, test failure, or unexpected behavior, before proposing fixes.
brainstorming
You MUST use this before any creative work - creating features, building components, adding functionality, or modifying behavior. Explores user intent, requirements and design before implementation.
auto-perf-optimize
Run agent-driven VS Code performance or memory investigations. Use when asked to launch Code OSS, automate a VS Code scenario, run the Chat memory smoke runner, capture renderer heap snapshots, take workflow screenshots, compare run summaries, or drive a repeatable scenario before heap-snapshot analysis.
chat-perf
Run chat perf benchmarks and memory leak checks against the local dev build or any published VS Code version. Use when investigating chat rendering regressions, validating perf-sensitive changes to chat UI, or checking for memory leaks in the chat response pipeline.
chat-pet-sprite-creation
Use when creating or changing VS Code chat pet sprite art, sprite sheets, state animations, eye treatments, Stable/Insiders variants, or pet transitions under src/vs/workbench/contrib/chat/browser/widget/media/chatPet.
cpu-profile-analysis
Analyze V8/Chrome CPU profiles (.cpuprofile) and DevTools trace files (Trace-.json). Use when: profiling performance, investigating slow functions, comparing code paths, finding bottlenecks, analyzing timeToRequest, understanding call trees from sampling profiler data, analyzing layout/paint/rendering, investigating…