Core

3 agenti · 0 workflow · 19 skills.

Agenti

• Specter — CISO / Quick Ops
• Chronicle — Security Documentation Specialist
• Referee — Exercise Referee

Skill

• spectra-init — Initialize SPECTRA project configuration and load config variables. Use when any skill needs module-specific configuration values, or when setting up a new SPECTRA project.
• spectra-help — Analyzes current state and user query to answer SPECTRA questions or recommend the next skill(s) to use. Use when user asks for help, spectra help, what to do next, or what to start with.
• spectra-war-room — Multi-agent adversarial debate. Use when the user says “war room” or wants Red vs Blue discussion on a security topic.
• spectra-debrief — Post-engagement or post-exercise review and lessons learned. Use when the user says “debrief” or “lessons learned” or “post-mortem”.
• spectra-new-engagement — Create a new security engagement with scope, rules of engagement, and authorization context. Use when starting a new pentest, assessment, or exercise.
• spectra-scope-check — Verify that a target or action is within the current engagement scope. Use before any offensive action.
• spectra-close-engagement — Close and archive a completed engagement with final reporting and evidence packaging.
• spectra-report-generator — Generate standardized security reports — pentest, incident, compliance, executive brief.
• spectra-evidence-chain — Manage chain of custody for digital evidence. Track acquisition, storage, and transfer.
• spectra-executive-brief — Generate C-level executive summary from engagement findings.
• spectra-duel-adjudication — Duel adjudication workflow. Use for Red/Blue ledger correlation, Referee scorecards, fairness review, detection latency, mitigation credit, and evidence-backed exercise outcomes.
• spectra-attack-path — Generate an evidence-anchored attack-path graph from engagement kill-chain and findings, with optional Blue detection overlay. Use to visualize how findings chain into impact and where detection gaps are.
• spectra-tool-run — Run a vetted security tool against an in-scope target, gated by engagement scope and Rules of Engagement. Use when an agent needs to actually execute recon/scanning tooling, not just plan it.
• spectra-exec-target — Run authorized commands on an operator-declared host over SSH, pinned to its host fingerprint and gated by engagement scope. Use to execute against a real in-scope target host.
• spectra-benchmark — Run a reproducible capability benchmark — gated tool runs against a known lab, scored against expected markers, with a deterministic grade. Use to prove SPECTRA finds what it claims.
• spectra-visual-evidence — Register a screenshot, diagram, or photo as hashed, custody-tracked evidence and anchor a visual analysis to it. Use when a finding rests on something you can see, not just text.
• spectra-new-module — Scaffold a new SPECTRA domain module (agent + workflow + config) with a valid structure. Includes an OT/ICS preset. Use to extend SPECTRA into a new domain.
• spectra-posture-diff — Snapshot an engagement’s security posture and diff two snapshots into a scored delta (improved/regressed). Use to re-assess the same scope over time and measure change — SPECTRA-native, not continuous monitoring.
• spectra-remediation-export — Export an engagement’s findings into remediation-ready formats — SARIF, CSV, or a Markdown ticket pack. Use to hand findings to engineering, CI, or a tracker. Deterministic export, not a live integration.