prd
Skill35 starsChanged yesterdayMITView on GitHub
Write or upgrade a Seekstone PRD from a Linear issue (e.g. SHA-290) or a feature idea, using docs/PRD_TEMPLATE.md and the repo's admission bar, write-safety and docs-sync gates. Use before implementing any new tool, prompt, resource, or user-facing behavior.
130 other projects use a version of this.
Adopted by
Most recently taken up by
- chaunsin/agent-skillstheir own version, first seen Nov 16, 2026
- ohvignas/lancetonsitetheir own version, first seen Nov 16, 2026
- librosdeia/builder-method-claude-plugintheir own version, first seen Nov 15, 2026
- realRoc/skillstheir own version, first seen Nov 15, 2026
- felipefontoura/claude-sdd-kittheir own version, first seen Nov 15, 2026
- felipeam86/pipe-skillstheir own version, first seen Nov 15, 2026
Used with
Projects with it also use context-optimize, usecase 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.
ibahgatSkillcontext-optimize
Improve context efficiency and cache stability for long or complex sessions.
Also in 3 projects that use this
gabriellstSkillusecase
Create an application use case. Use when implementing business operations like CreateOrder, ProcessPayment, CancelBooking. Use this skill for any command/mutation that orchestrates domain logic, whether simple CRUD or complex multi-step transactions with sagas. Dispatch hub — routes to typescript/go variants by file extension.
Also in 3 projects that use this
jidechaoSkillmodern-go
Modernize Go code by applying version-appropriate idioms and APIs (gofix-style transformations). Scans go.mod for the Go version, then transforms Go source files to use modern patterns—from Go 1.0 through 1.26+. Use when the user says "现代化","现代Go语言", "地道的", "idiomatic", "modernize", "modern-go", "update Go code", "gofix", or wants to upgrade Go idioms.
Also in 3 projects that use this
ulises-jeremiasSkilltrd
WHAT — Draft and review a Technical Requirements Document (TRD) using the template, typically
Also in 3 projects that use this
jidechaoSkillloop-it
Automated issue loop with checkpoint/resume: fetch open GitHub issues → dependency-aware topological sort → implement each issue end-to-end → review with /review-it → document with /note-it → ship with /ship-it → repeat. Persists state to .loop-state.json for crash recovery. Triggers on: loop-it, loop issues, auto implement, 批量实现, 循环实现, 实现所有issue, 恢复循环, resume loop.
Also in 3 projects that use this
snarktankSkillralph
Convert PRDs to prd.json format for the Ralph autonomous agent system. Use when you have an existing PRD and need to convert it to Ralph's JSON format. Triggers on: convert this prd, turn this into ralph format, create prd.json from this, ralph json.
Also in 16 projects that use this
- +2 more from jidechao
Other versions130 projects use a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 130
- github/awesome-copilot★ 41ksame name
- snarktank/ralph★ 22ksame name
- SethGammon/Citadel★ 927same name
- smallnest/autoresearch★ 571same name
- snarktank/compound-product★ 542same name
- CatCatUncle/openworkbuddy★ 413same name
- smallnest/goal-workflow★ 283same name
- OffchainLabs/arbitrum-docs★ 252same name
- julianromli/ai-skills★ 222same name
- golbin/agent-skills★ 214same name
120 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
We split this skill into the rules it gives an agent
9 rules. 0 are common, 5 are distinctive, and 4 are very specific to this project.
0
Common
Most of the 107 edited versions we read keep these.
5
Distinctive
This author's own choices. Worth a look.
4
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Very specific: line 18
Read the issue (get_issue), its comments (list_comments) and relations (blocks, blocked-by, duplicates).
- Distinctive: line 21
Keep everything already decided.
- Distinctive: line 32
Every claim about the codebase names a real file or symbol, and you check each one before writing it:
- Distinctive: line 53
§7 Surfaces table: Keep only the rows that apply.
- Distinctive: line 70
§12 Open decisions: Every real choice the implementer would otherwise make alone (naming, prefix vs exact matching, one tool or two), each …
- Very specific: line 85
Run the prose through humanizer:humanizer first.
- Distinctive: line 91
Leave the issue's workflow state alone unless the user asks to change it.
- Very specific: line 95
No marketing figures from memory.
- Very specific: line 97
No implementation in this step.
“Common” is measured: kept by at least 6 in 10 of the 107 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
/prd
Turns a Linear issue or a feature idea into an execution-ready PRD in the shape of
docs/PRD_TEMPLATE.md, then stores it as the
issue's description. The template is the format. This skill is the process for
filling it in so the result holds up against the repo's real gates.
1. Gather context
For a Linear issue ($ARGUMENTS looks like SHA-123):
- Read the issue (
get_issue), its comments (list_comments) and relations (blocks, blocked-by, duplicates). Read the parent epic too. Its description sets the strategic lens. - Keep everything already decided. Merged-duplicate notes, acceptance criteria and constraints in the old description carry over into the new PRD. Don't drop them.
- Find sibling PRDs this one depends on or feeds (for example
vault_statsshows counts thatfind_orphanscomputes) and link them in §13.
For a feature idea, search Linear for an existing issue first (list_issues
with query). If one exists, use it. If not, ask before creating one.
2. Ground it in the code
Every claim about the codebase names a real file or symbol, and you check each one before writing it:
- What already exists. Read
packages/server/src/tool-list.tsand the relevantpackages/server/src/tools/*.ts. The index likely already holds the data (link graph, tags, mtimes). Say which structure the tool reads from. - Overlap. Find the closest existing tool. If a parameter on it would do the job, the PRD proposes that instead of a new tool (CONTRIBUTING.md: "If no task shows a win, the capability probably belongs as a parameter on an existing tool").
- For write tools, read
docs/WRITE-SAFETY.mdand one existing write tool of the same shape (one file:patch_note.ts; several files:move_note.ts).
3. Draft against the template
Fill every section of docs/PRD_TEMPLATE.md. If a section doesn't apply, write
N/A: <reason>. These sections must not be thin:
- §5 Surface / contract. The exact zod-style input schema and the output shape, with a realistic example response. Keep descriptions short: every word in a tool description is schema tax, paid in every session.
- §7 Surfaces table. Keep only the rows that apply.
- §8 Acceptance criteria. Binary and testable. Include at least one error or edge case and one payload-size criterion.
- §10 Non-functional. A payload budget in bytes. Retrieval-path work holds p95 ≤ 30 ms and about 2 KB per payload. Write tools list every write-safety item from the template.
- §10b Speed budget. Seekstone's thesis is that it's the fastest Obsidian MCP server, so weigh every design choice against speed first. Fill in the budget table: warm p95 for the new call, added cost to the boot index build and to each watcher event, and "no regression" on the existing bench rows, each with the command the PR uses to prove it. Name the fastest design you found. If you chose something slower, say why. A design that slows an existing benchmark row doesn't ship; take it to the user as an open decision.
- §10a Admission bar. Required for any new tool, prompt, or resource: why
not a parameter, the
tasks.jsontask, the expected token win, and the estimated schema-tax cost. - §12 Open decisions. Every real choice the implementer would otherwise make alone (naming, prefix vs exact matching, one tool or two), each with options and a recommendation. Don't include questions the code already answers.
Header table: Status Draft, Owner the issue's assignee or Shaq, the issue ID,
its current Linear priority and milestone.
4. Review with the user
Show the full draft before writing it anywhere. List the open decisions separately so they can be answered in one pass. Revise until it's approved.
5. Publish
- Run the prose through
humanizer:humanizerfirst. Code, identifiers, numbers and commands stay exact. - Replace the Linear issue description with the approved PRD (
save_issue), set its Status row toApproved, and keep any "merged from SHA-XX" note at the end. - Prefix the title with
PRD:if it doesn't have it yet. - Leave the issue's workflow state alone unless the user asks to change it.
Don'ts
- No marketing figures from memory. Benchmark numbers come from
benchmarks.jsonor the committed baseline reports. - No implementation in this step.
/prdends at an approved PRD.
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.

