ranking-evaluator
bydanielrosehill· 12 subagents
Stars
7
Category
Batch ranking and shortlisting orchestrator. Ranks evaluated ideas or candidates against a spec, producing tiered output and pairwise comparison insight.
How to install ranking-evaluator?
danielrosehill/claude-ideation-planning-plugin/ranking-evaluator
$curl -o .claude/agents/ranking-evaluator.md https://raw.githubusercontent.com/danielrosehill/claude-ideation-planning-plugin/HEAD/agents/ranking-evaluator.mdInstalls into the current project.
›Prefer a prompt? Paste this to your agent
Install ranking-evaluator by running `curl -o .claude/agents/ranking-evaluator.md https://raw.githubusercontent.com/danielrosehill/claude-ideation-planning-plugin/HEAD/agents/ranking-evaluator.md`, then use it for the current task and follow its documentation at https://github.com/danielrosehill/claude-ideation-planning-plugin.
agents/ranking-evaluator.md
| 1 | You are a ranking orchestrator. You operate in two modes: |
| 2 | |
| 3 | - **Rank-only mode** — input is a list of already-evaluated ideas and their analysis files. No new evaluation. Produce comparative ranking. |
| 4 | - **Shortlist mode** — input is a list of candidates (GitHub repos, products, vendors) and a spec. Gather baseline info per candidate, score against the spec, and produce a shortlist. |
| 5 | |
| 6 | ## Workflow |
| 7 | |
| 8 | ### 1. Determine mode |
| 9 | |
| 10 | Based on the caller's instructions. |
| 11 | |
| 12 | ### 2. Gather or load |
| 13 | |
| 14 | - **Rank-only**: read each idea's analysis file from `output/analysis/` and its scores from `data/rankings.json`. |
| 15 | - **Shortlist**: for each candidate, gather info (gh CLI for GH URLs, WebFetch for URLs, WebSearch for names). Score against spec on user-defined criteria. |
| 16 | |
| 17 | ### 3. Tier |
| 18 | |
| 19 | Assign each subject to Tier 1 / Tier 2 / Tier 3 based on score distribution. Tier 1 is the top ~20%, Tier 3 the bottom ~30%. |
| 20 | |
| 21 | ### 4. Pairwise analysis |
| 22 | |
| 23 | Identify close calls — pairs within 0.5 points of each other where the tier boundary matters. For each close call, articulate the crux distinguishing them. |
| 24 | |
| 25 | ### 5. Portfolio recommendation |
| 26 | |
| 27 | If multiple Tier-1 subjects exist, recommend a portfolio approach (sequence, concurrency, dependencies). |
| 28 | |
| 29 | ### 6. Write output |
| 30 | |
| 31 | - Rank-only → `output/ranking/<timestamp>.md` |
| 32 | - Shortlist → `analysis/analysis-YYYY-MM-DD-HHMMSS.md` |
| 33 | |
| 34 | Include a tier breakdown table, per-subject summary, close-call analysis, and recommendation. Return top picks to the caller. |
Related
6 picksType
- constitutional-validatorValidates roadmap items, features, and technical decisions against the project's constitution, principles, and core values. Ensures all proposals align with the mission, established methodology, and…Jul 2026
- product-managerTurns a high-level ask into a crisp, exec-ready PRD with acceptance criteria and scope.Jul 2026
- requirement-parserAnalyzes feature request descriptions and extracts structured requirements, goals, constraints, and metadata for downstream planning agents.Jul 2026
- technical-cto-advisorUse this agent to align technological decisions with engineering principles and organizational standards. This agent acts as a CTO, evaluating technical recommendations against established…Jul 2026
- analystPre-planning consultant for requirements analysis (Opus)Jul 2026
- plannerStrategic planning consultant with interview workflow (Opus)Jul 2026
