jefflyt Mandate 2.1.1 Code Static Evaluation Skill
Mandate
- ID: 2.1.1
- Title: Authorization and Access Control
Mitigates
- LLM06 Excessive Agency
- ASI01 Agent Goal Hijack
- ASI03 Identity and Privilege Abuse
- MCP01 Tool Poisoning and Rug Pull Attacks
Inputs
- Repository root path.
- Tech stack/language map.
- Optional architecture notes if provided.
Workflow
- Scope relevant code surfaces tied to this mandate.
- Run mandate-specific static checks and trace source-to-sink/control paths.
- Extract exact evidence with file/line references.
- Classify each control as implemented, partial, or missing.
- Produce partial-confidence findings and list required runtime/process evidence for closure.
Mandate-Specific Static Method
- Enumerate privileged actions exposed by agents/tools.
- Extract permission scopes/roles from code, configs, and IaC files.
- Flag wildcard scopes (
*), admin roles, broad standing credentials, and missing per-action authorization checks. - Detect direct calls to sensitive systems without mediation layers.
- Trace whether privilege assignment is contextual (task-scoped) or static/global.
- Emit partial findings with file/line evidence and probable abuse path.
Output Contract
Return a single structured result with:
- mandate_id, mandate_title
- static_status: implemented|partial|missing
- confidence: low|medium|high (static-only confidence)
- evidence: list of file/line findings
- gaps: unresolved items requiring operational proof
- next_evidence_requests: runtime/process artifacts needed
Guardrails
- Avoid final compliance pass/fail using code alone.
- Mark assumptions explicitly.
- Prefer deterministic checks over heuristics when possible.