agent-runtime-validati…
Plans and evaluates bounded runtime tests of AI agents, tools, memory, hooks, approvals, and scope enforcement. Use when static…
Synthesizes supplied evidence-backed defensive assessment results and administrative receipts into a scope-bound control-evidence matrix that preserves lineage, freshness, contradictions, unknowns, and gaps. Use when consolidating defensive control evidence without scanning,
Claude Code
$ npx -y skills add zero-lab-ai/skills --skill defensive-control-evidence-synthesis --agent claude-code
Codex
$ npx skills add zero-lab-ai/skills --skill defensive-control-evidence-synthesis -a codex -y
Cursor
$ npx skills add zero-lab-ai/skills --skill defensive-control-evidence-synthesis -a cursor -y
Gemini CLI
$ npx skills add zero-lab-ai/skills --skill defensive-control-evidence-synthesis -a gemini-cli -y
GitHub Copilot
$ npx skills add zero-lab-ai/skills --skill defensive-control-evidence-synthesis -a github-copilot -y
OpenCode
$ npx skills add zero-lab-ai/skills --skill defensive-control-evidence-synthesis -a opencode -y
Windsurf
$ npx skills add zero-lab-ai/skills --skill defensive-control-evidence-synthesis -a windsurf -y
name: defensive-control-evidence-synthesis description: "Synthesizes supplied evidence-backed defensive assessment results and administrative receipts into a scope-bound control-evidence matrix that preserves lineage, freshness, contradictions, unknowns, and gaps. Use when consolidating defensive control evidence without scanning, scoring, certifying, or rewriting findings." allowed-tools: Read Grep Glob
Build and validate a closed matrix from already-supplied specialist artifacts and sanitized administrative, validation, or remediation-verification receipts.
Never contact a system, scan, sample, query, test, or execute. Reject raw logs, configurations, ACLs, credentials, tickets, certificates, packets, live identifiers, commands, queries, rules, test bodies, attack paths, and remediation procedures. Never create, rewrite, merge, close, or severity-rank findings. Never calculate a posture score, grade, maturity level, compliance result, certification, or exploitability rank.
1. Accept only exact passed specialist artifacts and sanitized administrative, validation, or remediation-verification receipts already supplied. 2. Stop on operational or restricted content. Do not quote rejected values. 3. Preserve producer, artifact format, artifact digest, review receipt, scope, policy, time, window, origin lineage, upstream class, polarity, limitations, applicability, and strength. 4. Report only generic diagnostic codes; never reflect restricted values. 5. Keep all ten policy dimensions in the exact scope-by-dimension denominator. 6. Do not strengthen, reverse, or reinterpret an upstream claim. Preserve or weaken it only when freshness or custody requires that result. 7. Cap administrative evidence at `documented` or `configured`. 8. Treat a review pass as artifact-structure evidence, not proof of no exposure or proof that a control works. 9. Keep telemetry, detection, containment, and recovery evidence independent. 10. Accept a no-finding claim only from exact, complete, current upstream coverage. 11. Deduplicate shared origin lineage before assessing independence. 12. Require one open contradiction for eligible opposing independent claims. 13. Use `not-applicable` only when source, claim, binding, and cell carry the same exact current policy reason. 14. Require one immutable finding reference for each current or stale, otherwise-valid exposure claim; do not edit the upstream finding. 15. Treat remediation occurrence as unverified until a later exact verification receipt independently binds the earlier occurrence, scope, and dimension. 16. Keep evidence gaps administrative. They are not technical findings. 17. Keep handoffs exact-bound, chronological, defensive, and non-authorizing. 18. Return exact state counts, contradiction IDs, gap IDs, and finding-reference IDs. Never produce a score.
State the matrix identifier, exact state counts, contradictions, evidence gaps, and immutable finding references. Add: “Internal synthesis consistency only; no live interaction, score, compliance determination, certification, or finding modification occurred.”
the supplied artifact contract, not the absence of findings.
observation or validation.
detection outcome.
Portable, evidence-bound security skills for authorized defensive assessment, penetration testing, red teaming, detection engineering, purple teaming, and engagement operations. The collection contains 56 skills in 10 installable plugins from zero-lab.ai.
Plans and evaluates bounded runtime tests of AI agents, tools, memory, hooks, approvals, and scope enforcement. Use when static…
Reviews an MCP server's tool surface for tool poisoning, over-broad parameters, missing descriptions, and privilege escalation…
Maps identity, authority, instruction, data, memory, and approval propagation across agents and agent-to-agent handoffs. Use when…
Assesses whether direct, indirect, multi-turn, retrieved, or tool-mediated instructions can override an AI system's intended…
Maps cloud identity and authorization paths across AWS, Azure, GCP, and hybrid trust boundaries, separating proven access from…
Tracks embedded-device and firmware assessment coverage across exact units, acquisition methods, debug interfaces, stored…
Use for authorized security assessment of REST, GraphQL, WebSocket, or SOAP APIs, including discovery, authentication,…
Extract and decrypt Windows DPAPI-protected secrets (Credential Manager, browser logins/cookies, Wi-Fi credentials, KeePass keys)…
Clean and finalize authorized natural-language text intended for readers by auditing suspicious invisible Unicode and rewriting…