Agent Claude Code
Round 2 verification — reviews brutalist decisions for false negatives and severity over-corrections.
Agent Claude Code
Round 2 verification — reviews brutalist decisions for false negatives and severity over-corrections.
Agent Claude Code
Round 1 verification — re-runs PoCs with maximum skepticism, checks severity inflation, filters non-bugs.
Agent Claude Code
Analyzes proven findings for credible impact chains that elevate severity.
Agent Claude Code
Runs bounded deep surface-discovery and produces compact attacksurface, deep-summary, and surface lead artifacts.
Agent Claude Code
Tests one attack surface for vulnerabilities — spawned per-surface with injected context from the orchestrator.
Agent Claude Code
CosmWasm smart-contract vulnerability evaluator — spawned per smartcontract surface with chainfamily=cosmwasm, scaffolds and runs cargo test with cw-multi-test against the direct public HTTPS CosmWasm REST ladder.
Agent Claude Code
EVM smart-contract vulnerability evaluator — spawned per smartcontract surface, scaffolds and runs Foundry tests against the direct public HTTPS RPC ladder.
Agent Claude Code
Non-recursive evaluator-fanout leaf — tests one MCP-issued (bugclass × auth) cell, writes durable claim/coverage/technique state, and returns only BOBCHILDCELLDONE.
Agent Claude Code
Spawn-capable per-surface evaluator — actuates the brain-owned childfanoutplan through Claude's one supported teammate-to-subagent edge, one child per (bugclass × auth) cell. Transition-blind: discovered cross-surface pivots ride discoveredpivots[] up to the orchestrator.
Agent Claude Code
Move (Aptos + Sui) smart-contract vulnerability evaluator — spawned per smartcontract surface with chainfamily in {aptos, sui}, scaffolds and runs aptos move test or sui move test against direct public HTTPS Move RPC/REST ladders.
Agent Claude Code
Provider-neutral physical-security evaluator — plans bounded coverage and records opaque evidence references without direct hardware, transport, or provider authority.
Agent Claude Code
Generic TaskGraph evaluator shell — executes Transition and Hypothesis nodes the orchestrator dispatched via bobpreparenode. Carries the union of evaluator-family tools; the dispatched brief's allowedtoolsfornode[] is the per-spawn constraint enforced by the X.6 mechanical verifier on bobfinalizenode.
Agent Claude Code
Substrate / ink! smart-contract vulnerability evaluator — spawned per smartcontract surface with chainfamily=substrate, scaffolds and runs cargo test on ink! contracts against the direct public HTTPS Substrate JSON-RPC ladder.
Agent Claude Code
SVM (Solana) smart-contract vulnerability evaluator — spawned per smartcontract surface with chainfamily=svm, scaffolds and runs Anchor tests against the direct public HTTPS Solana RPC ladder.
Agent Claude Code
Collects bounded pre-grade evidence packs for final reportable findings (HTTP via bobhttpscan; SC via family runners).
Agent Claude Code
Round 3 verification — re-runs only REPORTABLE findings with fresh requests as final confirmation.
Agent Claude Code
Scores verified findings on 5 axes and issues SUBMIT/HOLD/SKIP verdict.
Agent Claude Code
Generates submission-ready vulnerability report from verified and graded findings.
Agent Claude Code
Task-less smart-contract recon expander — resolves proxies/diamonds/roles/linked addresses per chain and returns producedsurfaces[]; writes scratch only, never holds record/promote/finalize.
Agent Claude Code
Runs bounded normal surface-discovery — subdomain enum, live hosts, archived/crawled URLs, nuclei, JS/JWT extraction — and produces attacksurface.json.
Agent Claude Code
Calls the MCP surface router after surface-discovery and reports the capability-pack summary.
Command Claude Code
Manage Hacker Bob egress profiles for this project.
Command Claude Code
Run or resume a Hacker Bob security evaluation.
Command Claude Code
Create a Hacker Bob post-release improvement bundle for the currently installed Bob version.