Clear-Capabilities/agentic-security

Clear-Capabilities/agentic-security

Agent

Subagents that hold Edit MUST follow the same write-confinement contract the MCP server enforces (scanner/src/mcp/tools.js). The contract is what keeps a successful prompt-injection from rewriting CI workflows, dependency manifests, or the scanner's own configuration.

Clear-Capabilities/agentic-security

Agent

Safely apply dead-code cleanup batches identified by /trim-dead-code. Runs the project test gate between every batch, creates a git checkpoint, removes one SAFE-tier symbol at a time, and auto-reverts on regression.

Clear-Capabilities/agentic-security

Agent

ADVISORY (unscored — see PRD F3.5). Per-component CLEAN/SUSPICIOUS/MALICIOUS verdict for third-party dependencies. Use after /security-sca surfaces packages and you need to decide whether a vulnerability is malware vs. ordinary CVE.

Clear-Capabilities/agentic-security

Agent

Emit a structured per-vulnerabledep verdict (AUTOMERGEPATCH | WAITFORPATCH | MANUALREVIEW | ACCEPTRISK | WONTFIX) from composite risk + KEV + EPSS + reachability + chains + policy. Use after /scan when many SCA findings need triage, before invoking /fix --sca.

Clear-Capabilities/agentic-security

Agent

Combine individual security findings into multi-step attack chains (e.g., IDOR + missing auth = account takeover). Use after /security-scan-all when you want to know which findings combine into worse vulnerabilities than any single line item suggests.

Clear-Capabilities/agentic-security

Agent

Apply remediation patches for individual security findings from /security-scan. Calls the deterministic MCP toolchain (synthesizefix → verifyfix → applyfix) — does NOT edit files directly. Reports back what the deterministic verifier observed.

Clear-Capabilities/agentic-security

Agent

Read route handlers and find business-logic flaws that pattern matchers miss — broken authorization tier checks, missing negative test cases, race conditions, state-machine bypasses, intent/implementation mismatches. Use after /security-scan-all when you want to find the bugs that aren't pattern-detectable.

Clear-Capabilities/agentic-security

Agent

Score the security materiality of a git diff (or a PR's changes). Separates routine refactors from architectural risk — auth removed, new endpoints, new prompts with user input, new shell calls, new IaC privilege grants. Use after a PR is opened or before merging a feature branch.

Clear-Capabilities/agentic-security

Agent

For ONE finding the scanner produced from the current working directory, build a PoC (proof-of-concept) input + regression test (framework-idiomatic), trace the data flow step-by-step, and emit 3-5 adversarial variants for confirmed TPs — or emit PROBABLEFP when a static blocker is found, REFUSED for out-of-tree…

Clear-Capabilities/agentic-security

Agent

Score, dedupe, and rank a list of security findings by risk. Produces a sorted, deduped list ready for human or AI consumption. Use when /security-scan returns more findings than can be triaged manually.