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 commands/zealynxsecurity/krait/krait-pocgit clone --depth 1 https://github.com/ZealynxSecurity/kraitWrote 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/zealynxsecurity/krait/krait-poc)<a href="https://agentmods.dev/commands/zealynxsecurity/krait/krait-poc"><img src="https://agentmods.dev/badge/commands/zealynxsecurity/krait/krait-poc.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 | $0.00000 | $0.00777 |
| Opus 5 | $0.00000 | $0.00388 |
| Sonnet 5 | $0.00000 | $0.00155 |
| Haiku 4.5 | $0.00000 | $0.00078 |
Grade A, and why
krait-poc 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 — 57 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Krait PoC — Foundry Exploit Proof-of-Concept
Write and run a valid Foundry proof-of-concept that proves (or disproves) a Solidity exploit by asserting the actual harm on a forked chain or against local source.
Usage
/krait-poc # Build a PoC for the finding(s) in .audit/, or ask what to prove
/krait-poc src/Vault.sol:142 # Build a PoC for a suspected bug at a location
/krait-poc reproduce <incident> # Reproduce a known on-chain hack from its post-mortem
/krait-poc triage findings.md # Batch: verify a LIST of findings → one verdict table
/krait-poc triage .audit/krait-findings.json
Single finding → follow the 8-step workflow. A list of findings (the triage form,
a .md/.json of findings, or more than one pasted) → read and follow
~/.claude/skills/krait-poc/references/batch-triage.md, which produces a consolidated
verdict table.
Instructions
Read and follow ~/.claude/skills/krait-poc/SKILL.md, then work through its 8-step
workflow, loading the referenced files under ~/.claude/skills/krait-poc/references/ only
as each step directs.
Non-negotiables:
- Assert HARM, not mechanism. A PoC that proves a function is callable or a state is reachable is not valid. It must assert who loses what — snapshot the loss-bearing balance/state before, run the attack, assert the delta.
- Every address and function signature must come from a source you read (deployed
source, ABI via
cast, or the in-scope repo). Never invent a selector. - Pin the fork block. An un-pinned fork is not reproducible.
- Use the forge MCP (
forge_build/forge_test) to compile and run, not raw shell. - Max 5 compile attempts, then fall back to
[CODE-TRACE]with the specific blocker. - Never weaken an assertion to force a pass. If the harm does not reproduce, that is
[POC-FAIL]— a real result. - Inability to PoC is not a refutation. Only
[POC-FAIL](a harm-asserting test that ran and did not reproduce) argues a finding is invalid. A finding that is un-PoC-able by nature (trusted-actor, off-chain, cross-chain) or un-buildable here keeps its severity as[CODE-TRACE]— never mark it "invalid" for lack of a PoC.
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 First seen · 57 lines · 0 tokens per session scan A fca95ccef7ba
krait-poc is a command published in the GitHub repository ZealynxSecurity/krait (22 stars, last pushed 25d ago), licensed MIT. It costs nothing until one of its globs matches a file; then it loads 777 tokens. 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 commands, from other repositories
audit
Full security audit of a Solidity/Vyper/Rust contract or directory. Runs the entire vuln-skills library and dispatches DeFi specialist subagents based on detected protocol type.
pre-deploy
Interactive pre-launch security checklist — walk the operational and code-safety gates and produce a final GO / NO-GO with each item PASS / FAIL / N-A.
audit-live
Audit a deployed contract on a live chain. Pulls verified source from the block explorer, optionally forks the chain for live-state simulation.
audit-multi-chain
Diff the on-chain configuration of one contract deployed across multiple chains — owner, oracle, fees, timelock, pause state, proxy impl — and flag the chain that drifted.
bounty-submit
Submit a finding to a bug-bounty platform (Immunefi, Cantina) automatically.
diff-audit
Diff the contract against a canonical reference implementation (OZ, Solady, Uniswap V3, etc.) and flag suspicious deltas.