bookend-reconcile
Merges disparate agent reports into a single, evidence-based readiness and complexity assessment.
Install
mkdir -p .claude/skills/bookend-reconcile && curl -L -o skill.zip "https://agentskills.codes/api/skills/download/18896" && unzip -o skill.zip -d .claude/skills/bookend-reconcile && rm skill.zipInstalls to .claude/skills/bookend-reconcile
Activation
This is the description your AI agent reads to decide when to run this skill — the better it matches your request, the more reliably it fires.
Merges bookend agent reports into revised readiness, complexity, and decomposition plan. Produces the final evidence-backed assessment consumed by sprint-architect-agent.Key capabilities
- →Collect reports from bookend agents
- →Revise readiness scores based on agent evidence
- →Revise complexity scores based on agent evidence
- →Aggregate and rank risk factors
- →Produce a Decomposition Plan
- →Generate architect guidance and challenger questions
How it works
The skill collects JSON reports from bookend agents, revises initial readiness and complexity scores using the evidence from these reports, aggregates risk factors, and then produces a final DecompositionPlan.
Inputs & outputs
When to use bookend-reconcile
- →Reconciling agent-led sprint readiness
- →Updating project complexity scores
- →Generating final decomposition plans
About this skill
bookend-reconcile - Agent Report Reconciliation
Merges structured outputs from bookend agents into a single assessment. Revises the initial heuristic scores with evidence-backed data and produces the final DecompositionPlan.
Input Schema
{
"reports_dir": "/path/to/reports/bookend",
"original_readiness": {},
"original_complexity": {},
"bookends_dir": "/path/to/go-autonom8"
}
Instructions
1. Collect Available Reports
Read all JSON files from reports_dir:
- cartography.json
- dependency-graph.json
- contracts.json
- quality-audit.json
- risk-assessment.json
Not all will exist — only agents that were spawned produce reports.
2. Revise Readiness
Starting from original_readiness.score:
- If cartography created artifacts: add their weight to score
- If quality-audit shows coverage data: adjust type_coverage and test_proxy
- Recalculate documentation_ok against threshold
3. Revise Complexity
For each dimension with agent evidence, replace heuristic score:
- dependency_depth <- dependency-graph-agent.complexity_contribution
- integration_density <- contract-scanner-agent.complexity_contribution
- historical_churn <- risk-assessment-agent.complexity_contribution
- type_coverage <- quality-audit-agent.complexity_contribution
- test_proxy <- quality-audit-agent.complexity_contribution
Recompute total, reclassify tier.
4. Aggregate Risk Factors
Collect all risk_factors from agent reports:
- Deduplicate
- Rank: critical > high > medium > low
- Cap at 10 most significant
5. Produce Decomposition Plan
Map revised tier to agent configuration. Generate architect_guidance from top risk factors. Generate challenger_questions from shared contracts and circular deps.
Output Format
{
"skill": "bookend-reconcile",
"status": "success",
"reconciled_readiness": {
"original_score": 85,
"revised_score": 78,
"documentation_ok": true,
"agents_contributed": ["quality-audit-agent", "cartography-agent"]
},
"reconciled_complexity": {
"original_score": 64,
"revised_score": 72,
"original_tier": "complex",
"revised_tier": "complex",
"dimension_revisions": []
},
"decomposition_plan": {
"tier": "complex",
"suggested_agents": 3,
"requires_cartography": false,
"risk_summary": "...",
"architect_guidance": [],
"challenger_questions": []
}
}
Decision Logic
All agent reports missing?
→ Return original scores unchanged, flag "no_agent_evidence"
Only cartography ran?
→ Revise readiness only, keep complexity unchanged
Multiple agents ran?
→ Full reconciliation with dimension replacement
Token Efficiency
- Reads JSON reports only, no filesystem scanning
- Sub-second execution
- Deterministic merge logic
When not to use it
- →When bookend agent reports are not available
- →When the user needs only initial heuristic scores without evidence-backed data
Limitations
- →Returns original scores unchanged if all agent reports are missing
- →Revises readiness only if only cartography ran, keeping complexity unchanged
- →Caps aggregated risk factors at 10 most significant
How it compares
This workflow integrates and reconciles multiple agent reports to produce an evidence-backed assessment, which is more complete than relying on individual agent outputs or initial heuristics.
Compared to similar skills
bookend-reconcile side by side with the closest alternatives in the catalog.
| Skill | Installs | Updated | Safety | Difficulty |
|---|---|---|---|---|
| bookend-reconcile (this skill) | 0 | 13d | No flags | Advanced |
| planning-with-files | 233 | 6mo | Review | Intermediate |
| ralph-plan | 14 | 6mo | No flags | Beginner |
| agent-planner | 3 | 5mo | No flags | Intermediate |
Try saying
Example prompts that trigger this skill in your AI assistant.
You might also like
planning-with-files
davila7
Implements Manus-style file-based planning for complex tasks. Creates task_plan.md, findings.md, and progress.md. Use when starting complex multi-step tasks, research projects, or any task requiring >5 tool calls.
ralph-plan
mastra-ai
Interactive planning assistant that helps create focused, well-structured ralph-loop commands through collaborative conversation
agent-planner
ruvnet
Agent skill for planner - invoke with $agent-planner
llm-council
am-will
Orchestrate a configurable, multi-member CLI planning council (Codex, Claude Code, Gemini, OpenCode, or custom) to produce independent implementation plans, anonymize and randomize them, then judge and merge into one final plan. Use when you need a robust, bias-resistant planning workflow, structured JSON outputs, retries, and failure handling across multiple CLI agents.
agent-queen-coordinator
ruvnet
Agent skill for queen-coordinator - invoke with $agent-queen-coordinator
swarm-planner
am-will
[EXPLICIT INVOCATION ONLY] Creates dependency-aware implementation plans optimized for parallel multi-agent execution.