How ramani-viswanathan/iem-pm works
IEM-PM is a Claude Code skill for evidence-based audits of project and PMO artifacts against standards supplied by the organization. Its principal workflow runs from baseline and charter preparation through evidence analysis, gap classification and root-origin tracing, scoring and synthesis, then deterministic validation and report generation. Python handles baseline inventory and the structured-output pipeline; Claude performs the audit judgment. README statements establish the broader workflow and capabilities; sampled source confirms selected script responsibilities and imports, but not every skill-stage transition.
Main entry points: SKILL.md Audit state machine · manifest_to_findings.py Findings canonicalizer · baseline.py Baseline inventory · render.py HTML and TXT reports
Components 18
Audit inputs
- Declared standards
skills/intelligence-engine/knowledge - Delivery artifacts
Audit - Baseline inventory
skills/intelligence-engine/scripts/baseline.py - Knowledge index
skills/intelligence-engine/scripts/derive_knowledge_index.py
Audit reasoning
- PMO data charter
skills/intelligence-engine/references/charter-spec.md - Audit state machine
skills/intelligence-engine/SKILL.md - Define and measure
skills/intelligence-engine/SKILL.md - Classify and trace
skills/intelligence-engine/SKILL.md - Score and synthesize
skills/intelligence-engine/SKILL.md - Audit manifest
skills/intelligence-engine/assets/AUDIT_MANIFEST_template.md
Deterministic pipeline
- Findings canonicalizer
skills/intelligence-engine/scripts/manifest_to_findings.py - Schema validator
skills/intelligence-engine/scripts/schema.py - Stage timing log
skills/intelligence-engine/scripts/timing_log.py
Audit outputs
- Canonical findings
skills/intelligence-engine/scripts/findings.schema.json - HTML and TXT reports
skills/intelligence-engine/scripts/render.py - Scope limitation notice
skills/intelligence-engine/scripts/render_scope_limitation.py - Report templates
skills/intelligence-engine/assets
Other components
- Audit operator
Connections 22
- Audit operator → Declared standards: provides baseline
README.md - Audit operator → Delivery artifacts: provides artifacts
README.md - Audit operator → Audit state machine: runs skill
README.md - Declared standards → Baseline inventory: inventoried by
skills/intelligence-engine/scripts/baseline.py - Baseline inventory → Knowledge index: derives index
skills/intelligence-engine/scripts/baseline.py - Baseline inventory → Knowledge index: scans directories
skills/intelligence-engine/scripts/derive_knowledge_index.py - Audit state machine → PMO data charter: establishes charter
README.md - Audit state machine → Define and measure: runs stages
README.md - Audit state machine → Classify and trace: runs stages
README.md - Audit state machine → Score and synthesize: runs stages
README.md - Delivery artifacts → Define and measure: examined against baseline
README.md - PMO data charter → Define and measure: defines data meaning
README.md - Score and synthesize → Audit manifest: writes manifest
README.md - Audit manifest → Findings canonicalizer: parsed into findings
skills/intelligence-engine/scripts/manifest_to_findings.py - Findings canonicalizer → Schema validator: validates findings
skills/intelligence-engine/scripts/manifest_to_findings.py - Findings canonicalizer → Stage timing log: logs stage
skills/intelligence-engine/scripts/manifest_to_findings.py - Findings canonicalizer → Canonical findings: produces JSON
README.md - Canonical findings → HTML and TXT reports: rendered as reports
README.md - HTML and TXT reports → Stage timing log: logs stage
skills/intelligence-engine/scripts/render.py - HTML and TXT reports → Report templates: uses template
README.md - Audit state machine → Scope limitation notice: may halt to notice
skills/intelligence-engine/scripts/render_scope_limitation.py - Scope limitation notice → Report templates: renders template
skills/intelligence-engine/scripts/render_scope_limitation.py