Skill: Documentation Readiness
Purpose
Evaluate whether project documentation is sufficient for reliable AI-assisted planning and alignment, then provide a minimal, prioritized fill plan to close critical gaps.
Core Objective
Primary Goal: Produce a documentation readiness report that quantifies evidence quality by layer and prescribes the smallest set of document actions needed to reach target readiness.
Success Criteria (ALL must be met):
-
✅ Layer coverage assessed: Goal, requirements, architecture, milestones, roadmap, and backlog layers are evaluated
-
✅ Readiness scored: Each layer is scored as strong , weak , or missing
-
✅ Gaps prioritized: Missing and weak layers are ranked by impact on delivery and alignment
-
✅ Minimal fill plan produced: Actionable steps identify what to create/update first and why
-
✅ Boundaries preserved: The skill does not rewrite product requirements or redesign architecture content
-
✅ Report persisted: Output document is written to agreed location
Acceptance Test: Can a teammate start improving project documentation immediately using only this report, without guessing what to write first?
Scope Boundaries
This skill handles:
-
Documentation inventory and layer mapping
-
Readiness scoring and confidence impact analysis
-
Gap prioritization and minimum-fill planning
-
Handoff recommendations to specialized skills
This skill does NOT handle:
-
Full requirements authoring from vague intent (use analyze-requirements )
-
Full design synthesis (use brainstorm-design )
-
Structural docs bootstrap from scratch templates (use bootstrap-project-documentation )
-
Post-task drift calibration (use execution-alignment )
Handoff point: After report delivery, hand off creation/update actions to the relevant documentation or planning skill.
Use Cases
-
Alignment confidence is low: execution-alignment reports weak evidence quality
-
New repo with partial docs: Team needs to know the minimum docs to add first
-
Pre-release governance: Verify documentation sufficiency before milestone closure
-
Documentation debt triage: Prioritize docs work without boiling the ocean
Behavior
Phase 0: Resolve Scope and Mapping
-
Resolve docs root and optional custom path mapping
-
Detect expected layer paths (template defaults first):
-
Goals: docs/project-overview/
-
Requirements: docs/requirements-planning/
-
Architecture: docs/architecture/
-
Milestones/Roadmap/Backlog: docs/process-management/
-
Accept backward-compatible aliases when present (for example, docs/requirements/ )
Phase 1: Inventory and Evidence Collection
-
Enumerate relevant docs per layer
-
Check freshness signals (last updated date, stale references, broken links where obvious)
-
Label evidence source:
-
canonical : project planning docs
-
secondary : issue/PR/commit context
-
none
Phase 2: Readiness Scoring
Assign one readiness level per layer:
-
strong: canonical docs exist and appear current enough for decision support
-
weak: docs exist but incomplete, stale, or dependent on secondary evidence
-
missing: no usable docs for the layer
Score overall readiness:
-
high : no missing layers and at most one weak layer
-
medium : one or more weak layers, no critical missing layer
-
low : at least one critical missing layer (requirements, architecture, or roadmap/backlog)
Phase 3: Gap Prioritization
For each gap, rate:
-
impact : high | medium | low
-
effort : small | medium | large
-
owner : suggested role/team
-
dueWindow : this-sprint | next-sprint | backlog
Prioritize by highest delivery risk reduction per smallest effort first.
Phase 4: Minimal Fill Plan
Produce the smallest set of actions needed to raise readiness to target level:
-
Layer to fix first and expected outcome
-
Exact document path(s) to create/update
-
Suggested handoff skill (bootstrap-project-documentation , analyze-requirements , brainstorm-design )
-
Stop condition for this cycle
Phase 5: Persist Report
Write to:
-
Default: docs/calibration/YYYY-MM-DD-doc-readiness.md
-
Or user-defined path
Input & Output
Input
-
Docs root path (default repository docs root)
-
Optional path mapping for non-template projects
-
Optional target readiness (medium or high )
Output
Documentation Readiness Report
Date: YYYY-MM-DD Overall Readiness: high | medium | low Target Readiness: medium | high
Layer Readiness
- Goal: strong | weak | missing
- Requirements: strong | weak | missing
- Architecture: strong | weak | missing
- Milestones: strong | weak | missing
- Roadmap: strong | weak | missing
- Backlog: strong | weak | missing
Gap Priority List
- Gap: ... Impact: high | medium | low Effort: small | medium | large Owner: ... DueWindow: this-sprint | next-sprint | backlog
Minimal Fill Plan
- Path: ... Why now: ... Handoff skill: ... Done condition: ...
Machine-Readable Summary
overallReadiness: "medium"
layers:
requirements: "missing"
architecture: "weak"
gaps:
- id: "gap-req-core"
impact: "high"
effort: "small"
owner: "product-owner"
dueWindow: "this-sprint"
Restrictions
Hard Boundaries
-
Do NOT fabricate non-existent docs
-
Do NOT rewrite product strategy, requirements, or architecture decisions in this skill
-
Do NOT prescribe a full documentation overhaul when a minimal plan is sufficient
-
Do NOT mark readiness as high when critical layers are missing
Skill Boundaries (Avoid Overlap)
Do NOT do these (other skills handle them):
-
Template bootstrap and structural initialization -> bootstrap-project-documentation
-
Requirement content development -> analyze-requirements
-
Architecture/design decision workflow -> brainstorm-design
-
Post-task drift and recalibration -> execution-alignment
When to stop and hand off:
-
If the primary gap is requirements quality -> hand off to analyze-requirements
-
If the primary gap is architecture clarity -> hand off to brainstorm-design
-
If docs skeleton is missing broadly -> hand off to bootstrap-project-documentation
Self-Check
Core Success Criteria (ALL must be met)
-
All planning layers assessed
-
Readiness per layer scored with rationale
-
Gaps prioritized by impact and effort
-
Minimal fill plan includes concrete paths and handoffs
-
Output persisted to agreed path
Process Quality Checks
-
Canonical vs secondary evidence clearly separated
-
Critical missing layers not hidden in aggregate score
-
Recommendations are minimal and sequencing-aware
-
Handoff skill names are valid and current
Acceptance Test
Can the team improve readiness in one sprint using only the top 3 actions?
If NO: reduce ambiguity and tighten prioritization.
If YES: report is complete.
Examples
Example 1: Missing Requirements Layer
-
Findings: goals and roadmap exist, requirements docs missing
-
Readiness: low
-
Plan: create docs/requirements-planning/core-v1.md first using analyze-requirements
Example 2: Weak Architecture Layer
-
Findings: architecture docs exist but stale and contradicted by recent ADRs
-
Readiness: medium
-
Plan: refresh architecture decision docs, then rerun execution-alignment
Example 3: Empty Docs Skeleton
-
Findings: no structured docs tree, only README exists
-
Readiness: low
-
Plan: hand off to bootstrap-project-documentation for baseline structure, then rerun this skill