/assess
Full SAIL V2 gap analysis of the AI agent/system in the current repo — component inventory, all 91 catalog risks dispositioned, prioritized recommendations. Writes SAIL-assessment.md.
$ npx -y skills add pillar-labs/sail-skill --agent claude-codeHow it fires
How this command gets triggered: by you, by Claude, or both.
- Fires itselfClaude auto-loads it when your prompt matches the work.
- You can call itInvoke it directly when you want it.
- Slash command
/assess
Context preview
What this command does when you run it.
Full SAIL V2 gap analysis of the AI agent/system in the current repo — component inventory, all 91 catalog risks dispositioned, prioritized recommendations. Writes SAIL-assessment.md.
Command definition
assess.mddescription: Full SAIL V2 gap analysis of the AI agent/system in the current repo — component inventory, all 91 catalog risks dispositioned, prioritized recommendations. Writes SAIL-assessment.md.
argument-hint: "[path or design doc — defaults to the current repo]"
Run the SAIL V2 **"Assess a system / gap analysis"** workflow immediately. Do NOT run the interactive intake and do NOT ask scoping questions.
**Target:** "$ARGUMENTS" if non-empty (a repo path or a design document); otherwise the current repository.
Follow the workflow exactly as specified in `${CLAUDE_PLUGIN_ROOT}/skills/sail/SKILL.md`, section **"Assess a system / gap analysis"**:
1. Read `${CLAUDE_PLUGIN_ROOT}/skills/sail/references/risk-index.md` first — it is the map. 2. Extract the component inventory from the target yourself (agent-framework imports, agent/tool configs, MCP configs, system prompts, credential handling, schedules) and map it onto the 29-component taxonomy in `${CLAUDE_PLUGIN_ROOT}/skills/sail/references/definitions.md`. 3. Chart the components: a Mermaid diagram grouped by zone with trust boundaries, plus the element → SAIL component → layer → zone table. 4. Sweep all 91 risks in the index and give every one a disposition — Gap / Partial / Addressed / Verify / N/A. Read the full phase file for every risk dispositioned Gap, Partial, or Verify so mitigations are quoted from the catalog. 5. Deliver per the workflow's document structure, citing SAIL IDs everywhere. If the target is a git repository, write the deliverable to `SAIL-assessment.md` at its root; otherwise emit the document in the conversation.
If the target contains no agent artifacts, say so, assess what does exist (LLM app, pipeline assets), and mark absent components N/A — still disposition all 91 risks.
Read more
description: Full SAIL V2 gap analysis of the AI agent/system in the current repo — component inventory, all 91 catalog risks dispositioned, prioritized recommendations. Writes SAIL-assessment.md. argument-hint: "[path or design doc — defaults to the current repo]"
Run the SAIL V2 **"Assess a system / gap analysis"** workflow immediately. Do NOT run the interactive intake and do NOT ask scoping questions.
**Target:** "$ARGUMENTS" if non-empty (a repo path or a design document); otherwise the current repository.
Follow the workflow exactly as specified in `${CLAUDE_PLUGIN_ROOT}/skills/sail/SKILL.md`, section **"Assess a system / gap analysis"**:
1. Read `${CLAUDE_PLUGIN_ROOT}/skills/sail/references/risk-index.md` first — it is the map. 2. Extract the component inventory from the target yourself (agent-framework imports, agent/tool configs, MCP configs, system prompts, credential handling, schedules) and map it onto the 29-component taxonomy in `${CLAUDE_PLUGIN_ROOT}/skills/sail/references/definitions.md`. 3. Chart the components: a Mermaid diagram grouped by zone with trust boundaries, plus the element → SAIL component → layer → zone table. 4. Sweep all 91 risks in the index and give every one a disposition — Gap / Partial / Addressed / Verify / N/A. Read the full phase file for every risk dispositioned Gap, Partial, or Verify so mitigations are quoted from the catalog. 5. Deliver per the workflow's document structure, citing SAIL IDs everywhere. If the target is a git repository, write the deliverable to `SAIL-assessment.md` at its root; otherwise emit the document in the conversation.
If the target contains no agent artifacts, say so, assess what does exist (LLM app, pipeline assets), and mark absent components N/A — still disposition all 91 risks.
An agent skill that applies SAIL V2 — Pillar Security's Secure AI Lifecycle framework — to real systems.
Repo: pillar-labs/sail-skill
Other commands on sail-skill.
- /comply
SAIL V2 compliance checklist — ISO/IEC 42001, EU AI Act, OWASP LLM/Agentic, DASF, AIUC-1. All five frameworks by default, or only those named.
Open command - /rfp
Vendor RFP questionnaire from the SAIL V2 catalog — security-vendor coverage questionnaire by default, platform-vendor framing on request.
Open command