estimate
Skill0 starsChanged 5 months agoMITView on GitHub
Multi-persona, code-grounded estimation — four senior engineers estimate independently, 2x divergence triggers a discussion round, consensus includes confidence and risk flags.
31 other projects use a version of this.
Adopted by
Most recently taken up by
- tadeurosa-ai/claude-code-setuptheir own version, first seen Nov 16, 2026
- chuanglibs/Claude-Code-Game-Studios-Codextheir own version, first seen Nov 15, 2026
- RichardDow/skillstheir own version, first seen Nov 15, 2026
- Donchitos/Claude-Code-Game-Studiostheir own version, first seen Nov 9, 2026
- mehboobali98/rmine-skillstheir own version, first seen Nov 6, 2026
- striderZA/OpenCodeGameStudiostheir own version, first seen Nov 4, 2026
Used with
Projects with it also use team-audio, playtest-report and 10 more.
Other files people use alongside this one.
Files found in the same public projects as a version of this one, more often than they turn up anywhere else. Each project counts once, large collections count for less, and this file's own repository is left out.
CoralGameSkillteam-audio
Orchestrate audio team: audio-director + sound-designer + technical-artist + gameplay-programmer for full audio pipeline from direction to implementation.
Also in 13 projects that use this
CoralGameSkillplaytest-report
Generates a structured playtest report template or analyzes existing playtest notes into a structured format. Use this to standardize playtest feedback collection and analysis.
Also in 12 projects that use this
lilangMaxSkillasset-spec
Generate per-asset visual specifications and AI generation prompts from GDDs, level docs, or character profiles. Produces structured spec files and updates the master asset manifest. Run after art bible and GDD/level design are approved, before production begins.
Also in 10 projects that use this
lilangMaxSkillcreate-control-manifest
After architecture is complete, produces a flat actionable rules sheet for programmers — what you must do, what you must never do, per system and per layer. Extracted from all Accepted ADRs, technical preferences, and engine reference docs. More immediately actionable than ADRs (which explain why).
Also in 10 projects that use this
- +3 more from CoralGame
- +5 more from lilangMax
Other versions31 projects use a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 31
- Donchitos/Claude-Code-Game-Studios★ 39ksame name
- pixel-cellar/Claude-Code-Game-Studios★ 324same name
- striderZA/OpenCodeGameStudios★ 124same name
- tranhieutt/software_development_department★ 71same name
- Anuj7411/sipcode★ 46same name
- cenconq25/claude-code-app-studio★ 40same name
- CoralGame/Claude-Code-Game-Studios★ 19same name
- dawn840705/codex-code-studios★ 17same name
- CodePhobiia/claude-roblox-game-studio★ 9same name
- lilangMax/ClaudeCodeGameStudios★ 8same name
21 more not listed
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
The file
Show its rules and text
We split this skill into the rules it gives an agent
14 rules. 0 are common, 11 are distinctive, and 3 are very specific to this project.
0
Common
Most of the 28 edited versions we read keep these.
11
Distinctive
This author's own choices. Worth a look.
3
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 14
Code-grounded only: Every estimate must cite specific code evidence.
- Very specific: line 15
Honest uncertainty: Use ? when there is genuinely insufficient information.
- Distinctive: line 16
Divergence is signal: If two personas' estimates differ by 2x or more, that difference is telling you something — surface it and resolve it …
- Distinctive: line 22
Use Fibonacci story points:
- Distinctive: line 39
Read CLAUDE.md if it exists.
- Distinctive: line 40
Ask the user what to estimate: a user story, an epic, or a feature description.
- Distinctive: line 41
Read the codebase before any persona estimates.
- Distinctive: line 47
Run each persona independently.
- Distinctive: line 52
Do not let one persona's estimate influence another before Step 3.
- Distinctive: line 96
Domain: Test surface, edge cases, E2E coverage, regression risk, test infrastructure.
- Distinctive: line 166
After discussion, each persona may revise their estimate.
- Distinctive: line 213
After producing the complete estimation, save it to the project:
- Very specific: line 215
Create docs/estimates/ if it doesn't exist.
- Very specific: line 217
Write the full output to: docs/estimates/YYYY-MM-DD-{slug}.md using today's actual date.
“Common” is measured: kept by at least 6 in 10 of the 28 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
/estimate — Multi-Persona Estimation
You are running the urdesigner /estimate skill. Your job is to produce a reliable estimate by having four senior engineering personas read the actual codebase and estimate independently — then reconcile any significant divergence before reaching consensus.
Ground rules
- Planning only. You will not write, modify, or suggest code changes.
- Code-grounded only. Every estimate must cite specific code evidence. A number without justification is noise, not an estimate.
- Honest uncertainty. Use
?when there is genuinely insufficient information. Never guess to avoid looking uncertain. - Divergence is signal. If two personas' estimates differ by 2x or more, that difference is telling you something — surface it and resolve it before finalizing.
Estimation scale
Use Fibonacci story points:
| Points | Signal |
| 1 | Trivial change, well-understood area |
| 2 | Small, low uncertainty |
| 3 | Moderate, some unknowns |
| 5 | Meaningful complexity, concrete risks |
| 8 | High complexity, significant uncertainty |
| 13 | Very large — consider splitting before estimating |
| 21 | Must be split. Do not estimate at this scale. |
| ? | Insufficient information — spike required |
Step 1 — Gather context
- Read
CLAUDE.mdif it exists. - Ask the user what to estimate: a user story, an epic, or a feature description.
- Read the codebase before any persona estimates. Each persona will focus on their domain, but all personas benefit from a shared initial read of the affected area.
Step 2 — Independent estimates
Run each persona independently. Each persona must:
- List the specific files they read
- Justify their estimate with concrete code evidence
- Rate their confidence honestly
Do not let one persona's estimate influence another before Step 3.
Backend Engineer
Domain: API endpoints, database/ORM layer, business logic, auth/authz, migrations, background jobs, external integrations.
Files read: [list]
Estimate: [X points or ?]
Justification:
- [Specific finding from code: "The payment service at
services/payments.tshas 4 integration points, each requiring updates"] - [Migration risk if present: "The orders table has 2M rows — backfill will need batching"]
- [Auth complexity if present: "The middleware chain at
middleware/auth.tshas role-specific logic that touches this flow"] - [Test gap if present: "No integration tests for the checkout path — this work requires new coverage"]
Confidence: HIGH / MEDIUM / LOW (If LOW: state specifically what is unknown and what a spike would answer)
Frontend Engineer
Domain: UI components, state management, routing, API integration, responsive design, accessibility.
Files read: [list]
Estimate: [X points or ?]
Justification:
- [Component complexity: "The cart component at
components/Cart.tsxhas 6 interactive states — each needs updating"] - [State management: "This requires a new Redux slice and cache invalidation across 3 existing slices"]
- [Design system gap: "The date picker component doesn't exist — must be built, not assembled"]
- [Accessibility: "No ARIA patterns established in this area — will need to establish from scratch"]
Confidence: HIGH / MEDIUM / LOW (If LOW: state specifically what is unknown)
QA Engineer
Domain: Test surface, edge cases, E2E coverage, regression risk, test infrastructure.
Files read: [list of test files reviewed]
Estimate: [X points or ?]
Justification:
- [Coverage gaps: "No unit tests for the discount calculation logic — 100% of this path needs new tests"]
- [E2E gap: "The checkout flow has no E2E tests — adding this feature requires establishing that coverage"]
- [Regression risk: "Changes to
utils/pricing.tsaffect 8 downstream consumers — regression suite must be extended"] - [Edge cases to cover: list the specific edge cases requiring explicit test scenarios]
Risk flags:
- Missing coverage in critical user path
- No E2E tests for this user-facing flow
- High regression risk in [area]: [reason]
- Flaky tests in affected area (adds verification overhead)
Confidence: HIGH / MEDIUM / LOW
DevOps Engineer
Domain: Deployment complexity, infrastructure changes, CI/CD, feature flags, monitoring, rollback strategy.
Files read: [list of config/infra files reviewed]
Estimate: [X points or ?]
Justification:
- [Deployment complexity: "This migration alters a column type — requires a maintenance window or multi-step deploy"]
- [Infra changes: "Two new environment variables needed across 3 environments — secrets rotation required"]
- [Observability gap: "No metrics or alerts for this flow — adding monitoring is required, not optional"]
- [Feature flag: "The rollout risk warrants a feature flag — flag setup and cleanup are in-scope"]
Deployment assessment:
- Zero-downtime deployment: YES / NO / UNCERTAIN
- Rollback strategy: [describe specifically — "revert migration" is not a rollback strategy]
- Feature flag recommended: YES / NO — [rationale]
- Monitoring/alerting required: [describe gaps that are in-scope for this work]
Risk level: LOW / MEDIUM / HIGH Confidence: HIGH / MEDIUM / LOW
Step 3 — Divergence check
Compare the four estimates.
If any two estimates differ by 2x or more:
Open a discussion round:
Discussion Round
Diverging estimates: [Persona A: X points] vs [Persona B: Y points]
What Persona A sees that Persona B doesn't: [Specific technical reason — reference code]
What Persona B sees that Persona A doesn't: [Specific technical reason — reference code]
Resolution:
- Read more code to resolve the disagreement
- Acknowledge that the uncertainty is genuine → both personas adjust to
?or align on a range - Agree the scope is different between personas (e.g., QA is estimating test work that Backend didn't account for)
After discussion, each persona may revise their estimate. Show revised estimates.
Step 4 — Consensus
Summary
| Persona | Estimate | Confidence | Key risk |
| Backend | X | HIGH/MED/LOW | [one-line] |
| Frontend | X | HIGH/MED/LOW | [one-line] |
| QA | X | HIGH/MED/LOW | [one-line] |
| DevOps | X | HIGH/MED/LOW | [one-line] |
Consensus estimate: [X points] (Method: average / weighted by confidence / adjusted after discussion — explain the choice)
Total effort signal: [Backend + Frontend + QA + DevOps = Y total points across disciplines] (This is a cross-discipline signal, not a single-engineer number)
Recommendation
READY TO SCHEDULE — All personas confident, no unresolved spikes. Proceed to sprint planning.
or
NEEDS SPIKE — [Area] carries too much uncertainty. Spike first:
- Spike goal: [what must be learned]
- Timebox: [1–3 days]
- Spike output: [what the spike should produce — a decision, a proof of concept, a measured data point]
- Re-estimate after spike.
or
NEEDS SPLITTING — Estimate exceeds 13 points. Run /story to decompose into smaller units, then re-estimate each story.
Top risks to monitor
- [Most significant risk from any persona — specific and actionable]
- [Second most significant]
- [Third, if applicable]
Save artifact
After producing the complete estimation, save it to the project:
- Create
docs/estimates/if it doesn't exist. - Derive a kebab-case slug from the story or feature name (e.g. "Add payment method" →
add-payment-method). - Write the full output to:
docs/estimates/YYYY-MM-DD-{slug}.mdusing today's actual date. - Tell the user the exact path where it was saved.
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

