audithub-skills

Generate best-effort DeFi Vanguard custom detectors from English security invariants for Solidity projects. Use when the user provides invariant lists, properties, protocol rules, assumptions, or expected behaviors and wants them converted into PAQL or Detector.register Luau detectors, usually one detector per invariant plus an optional detector config. This skill should inspect target contracts and write detector definitions when requested, but must not run Vanguard to validate them.

Use for AuditHub DeFi Vanguard workflows such as preparing static analysis runs, interpreting user-provided DeFi Vanguard results, and coordinating triage for Solidity projects. Trigger when a user asks for DeFi Vanguard analysis preparation, vulnerability triage based on DeFi Vanguard output, or static analysis findings.

Discover security-relevant invariants for smart-contract protocols. Use when the agent is asked to read Solidity, EVM, DeFi, token, vault, bridge, governance, staking, liquidation, oracle, or protocol code and produce a complete list of invariants whose violation is a viable attack surface causing loss, theft, denial of service, privilege escalation, accounting corruption, unfair value extraction, or harm to users/protocols. Also use when asked to improve, audit, or generalize invariant lists for fuzzing, formal verification, specs, or security review without writing exploits.

Backward-compatible orchestrator for AuditHub OrCa fuzzing workflows. Use when a user asks for a full or multi-stage OrCa campaign, autonomous OrCa setup/run/tuning, target selection, live-state setup, local deployment setup, smoke-run debugging, call_metrics.json analysis, setup tuning, property discovery, [V] specs, hints, counterexample triage, or final OrCa reporting for Solidity projects.

Coordinate full or multi-stage AuditHub OrCa fuzzing campaigns for Solidity projects. Use when a user asks to run an autonomous OrCa campaign, plan and execute setup/tuning/spec/long-run loops, resume an OrCa campaign, or coordinate target selection, deployment setup, smoke runs, call_metrics.json analysis, tuning, [V] specs, counterexample triage, and final reporting.

Prepare live-state OrCa campaign inputs for deployed Solidity contracts. Use when a campaign will fuzz on-chain state, needs chain/block/address/name/ABI data, needs Etherscan or explorer ABI handling, needs proxy implementation ABI resolution, or needs on-chain deployment JSON validation.

Translate English protocol properties into OrCa [V] specifications and validation notes. Use when a user asks for [V] specs, wants properties.md converted into OrCa specs, needs invalid [V] specs repaired from call_metrics.json spec errors, or needs weakened/strengthened/related [V] variants documented.

Analyze OrCa call_metrics.json outputs for fuzzing efficiency and blockers. Use when a user provides call_metrics.json, asks why functions revert or have low successful-call percentages, wants hint usage analysis, needs per-function metrics triage, or wants tuning hypotheses for an OrCa campaign.

Analyze OrCa counterexamples and classify real bugs versus specification issues. Use when OrCa reports violated [V] specs, a user provides counterexample sequences, traces, or failing artifacts, or when a campaign needs PoC guidance, Foundry reproduction planning, exploitability assessment, and write-up material.

Summarize final OrCa campaign results for Solidity protocols. Use after long OrCa runs or campaign completion to report confirmed findings, counterexample status, interesting call_metrics.json metrics, fuzzing efficiency, coverage gaps, residual risks, and recommended follow-up.

Write OrCa hints from function execution preconditions and metrics blockers. Use when functions have low success due to complex argument values, required state, balances, approvals, access control, or helper-derived values, or when call_metrics.json hint usage shows invalid or failing hints.

Build or update realistic local deployment scripts for OrCa fuzzing campaigns. Use when live-state fuzzing is unavailable or unsuitable, when a Foundry or Hardhat protocol needs an OrCa deployment script, or when setup needs mocks, initialized contracts, roles, balances, approvals, liquidity, oracle values, or default fuzz users.

Discover English-language protocol properties for OrCa fuzzing. Use when a user wants real specifications after setup tuning, asks what invariants or temporal properties should hold, needs code/tests/docs reviewed for OrCa properties, or wants a properties.md handoff before [V] translation.

Tune OrCa campaign setup after metrics analysis. Use when an OrCa run has poorly performing functions, access-control/balance/approval/state blockers, failing hints, missing target contracts, or when the campaign needs users, hints, helper contracts, deployment-state changes, fuzz targets, or blacklists while preserving fuzzing breadth.

Create dummy OrCa specs, perform or inspect short initial OrCa smoke runs, and debug setup failures. Use when a campaign needs the universal dummy spec, a first 1-2 minute run, diagnosis of failed local deployment scripts, diagnosis of malformed on-chain deployment JSON, or run-ledger updates for early OrCa attempts.

Select OrCa fuzzing targets and choose live-state vs local-deployment mode for Solidity protocols. Use when a user provides no explicit target list, asks which contracts to fuzz, wants critical contract selection for a bug bounty or open-ended review, or needs dependencies and mode recommendations for an OrCa campaign.