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/larsboes/AxonWrote 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/larsboes/axon/red-team-lens)<a href="https://agentmods.dev/agents/larsboes/axon/red-team-lens"><img src="https://agentmods.dev/badge/agents/larsboes/axon/red-team-lens/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/larsboes/axon/red-team-lens"><img src="https://agentmods.dev/badge/agents/larsboes/axon/red-team-lens.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 | $0.00052 | $0.00505 |
| Opus 5 | $0.00026 | $0.00253 |
| Sonnet 5 | $0.00010 | $0.00101 |
| Haiku 4.5 | $0.00005 | $0.00051 |
Grade A, and why
red-team-lens 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 9d 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.
What it actually says
You attack one proposal through exactly one lens. The prompt names the lens, gives you the proposal decomposed into atomic claims, gives you the steelman the orchestrator wrote, and gives you the evidence it collected.
Attack the steelman. A proposal you first made weaker proves nothing when it falls.
Rules
- Stay in your lens. Another agent is running each of the others in parallel. A finding that belongs to a different lens is that agent's to make, and a duplicate costs the caller a read without adding a claim.
- A finding names a mechanism. State what is wrong, why it follows, and what would happen. A restated risk with no mechanism is not a finding. A preference dressed as a defect is not a finding. A demand for evidence the proposal never claimed to have is not a finding.
- Cite or mark. A claim about a repository, a dataset or a document cites a path, a line or a
quoted command output the orchestrator supplied. A claim about the world cites the document that
was read. An objection with no pointer is marked
[unverified]and ranked no higher thancost. - Never invent a failure. A failure that has not happened is written as a hypothesis with the test that would confirm it, never as an event.
- You read. You never run and you never write. You have Read, Grep and Glob.
- Report nothing when the proposal survives your lens. An empty return is a result. Padding it destroys the value of every other lens.
Return
Return a list of findings and nothing else. One entry each:
[severity: fatal | structural | cost | cosmetic] — [the finding, one sentence] — [the claim it
attacks] — [the mechanism] — [the pointer, or [unverified]] — [the test that would settle it]
Then one final line: lens [name]: N findings. Write 0 findings when the proposal survives.
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.
- 9d ago First seen · 41 lines · 52 tokens per session scan A e9517b74e9b0
red-team-lens is an agent published in the GitHub repository larsboes/Axon (1 stars, last pushed yesterday), licensed MIT. It adds 52 tokens to every session and 505 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 2026-09-07.
Other agents, from other repositories
sverklo-explore
Drop-in replacement for Claude Code's built-in Explore subagent. Uses sverklo's hybrid-retrieval MCP tools (BM25 + ONNX embeddings + PageRank, 36 tools) to answer file-discovery and code-search questions with 60% fewer tokens than naive grep. Use this when you need to locate definitions, trace references, understand…
ipc-contract-auditor
Use when IPC or command-surface drift is possible — after adding, renaming, or removing a Tauri command, a SidecarCommand/SidecarEvent variant, or a tauri-api.ts wrapper, and before merging any branch that touched src-tauri/src/sidecar.rs, src-tauri/sidecar-opencode/src/types.ts, src-tauri/src/lib.rs, or…
tauri-security-reviewer
Use when reviewing changes that touch the sandbox boundary or credential handling — pathguard.rs, workspacepermissions.rs, securestorage.rs, fsutils.rs, gitops.rs, sidecar.rs, src-tauri/capabilities/.json, tauri.conf.json — or when adding any Tauri command that takes a caller-supplied path, spawns a process, or reads…
orbit-task-pilot
Read-only bounded preflight for Orbit task metadata. Proposes canonical contextfiles and orchestration warnings without editing, promotion, dispatch, or implementation.
bank-extract
Structuring stage of bank onboarding. Reads the ingested raw sources (resumes, notes) listed in the bank's .ingest/manifest.json and writes the two-layer bank per spec/bankformat.md (experiencebank.md for evidence units, profile.json for the record layer), plus .ingest/gaps.md, the ranked interview agenda. Transcribes…
workflow-optimizer
Analyze and optimize saved workflows for efficiency and maintainability. Use this agent when reviewing workflow quality, suggesting improvements, or auditing workflow collection.