agentleFS
dougcunhadougcunha/TokenHound/.agents/skills/sdd-create-techspec/SKILL.md

sdd-create-techspec

Skill3 starsChanged 35 days agoMITView on GitHub

SDD TechSpec when a PRD exists and the solution must be specified; does not create requirements or a task plan.

The file

We split this skill into the rules it gives an agent

5 rules. About 0 are common, 4 are distinctive, and 1 is very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈4

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 1

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Output: every obligation has a decision or pending item.

    line 13
  2. Very specific:

    Read assets/techspec.template.md in full when drafting.

    line 17
  3. Distinctive:

    Output: all obligations covered; unresolved decisions explicitly pending, without imposing a percentage of coverage.

    line 18
  4. Distinctive:

    Write only the TechSpec. Report decisions, gaps, and tasks invalidated by the update.

    line 19
  5. Distinctive:

    Keep stable sources ahead of recovered code and state.

    line 22

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Create SDD TechSpec

  1. Resolve tasks/prd-[slug]/prd.md and techspec.md. Require the PRD; if missing, point to sdd-create-prd. Read the PRD once per version. Reuse an existing TechSpec; update only when authorized, preserving IDs. Output: exact sources and destination, with no implicit overwrite.
  2. Map every PRD obligation to a technical consequence. Inspect only involved modules, callers, contracts, persistence, errors, tests, and configuration. Other features' PRD, TechSpec, and tasks do not describe the current code: inspect the code and do not edit them to reflect this solution. Reuse existing patterns; justify new dependencies and components with a proven gap. Consult primary documentation for external technical questions. Measure the terrain where the change will land: read references/preparatory-refactoring.md in full and apply its measures to the files the feature will modify. The result is the baseline of pre-existing hits and, when structural debt and contact coincide, a preparatory refactoring recommendation with minimal scope. Existing code the feature only reads generates no recommendation. Output: every obligation has a decision or pending item; every component has a path, responsibility, and integration; every target file has a measured baseline and a destination for the debt found.
  3. Identify the stack per affected project. For C#/.NET, read references/dotnet.md in full before specifying validation. In desktop .NET, omit E2E execution and record E2E: omitted by desktop .NET policy; preserve acceptance with smaller tests and a manual script when needed. Also specify the quality profile: read references/quality-dotnet.md in full and select the rules this feature can violate, with class and verification command. A rule outside the profile is checked by no gate. An already decided deviation becomes DEC-NN and stops being a finding. Output: validation profile with evidence of stack, runner, projects, commands, and limitations; quality profile with relevant rules, classes, commands, and prior justifications.
  4. Read assets/techspec.template.md in full when drafting. Use DEC-01, CMP-01, TC-01, and PRD IDs. Cover applicable contracts, errors, edges, security, concurrency, rollback, and rollout without duplicating requirements. Every obligation must have a test or other proportional evidence; remove inapplicable sections. Output: all obligations covered; unresolved decisions explicitly pending, without imposing a percentage of coverage.
  5. Write only the TechSpec. Report decisions, gaps, and tasks invalidated by the update. In the orchestrated flow, return it for plan preparation and joint technical HIL. Output: reviewable artifact with identified impacts; conflicts between PRD/code/contract that require human decisions were not invented.

Keep stable sources ahead of recovered code and state; reread only changed versions. Reading order helps consistency but does not guarantee provider caching.

Download now (ZIP)

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
What did you use it for? e.g. Reviewing pull requests in a Rails app

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.