Docs Interfaces
jh941213/my-cc-harness
Generate interface/API docs — OpenAPI 3.1/AsyncAPI 3.0 specs, API topology diagrams, interface flow (sequence) diagrams, API changelog.
A skill your agent uses to derive the API contract for a feature — an OpenAPI 3.1 document at docs/features/{slug}/contracts/openapi.yaml plus a drift/sync report (and an events doc when the feature…
$ npx skills add genkovich/sdd --skill api -a claude-codeProject install by default; add -g for ~/.claude/skills/.
$ gh skill install genkovich/sdd api --agent claude-codeProject scope by default; add --scope user for a personal install. Needs GitHub CLI 2.90.0 or later (public preview).
$ git clone --depth 1 https://github.com/genkovich/sdd.git skills-src && mkdir -p .claude/skills && cp -r skills-src/skills/api .claude/skills/api && rm -rf skills-srcUse ~/.claude/skills/ instead of .claude/skills for a personal install. The folder must contain SKILL.md.
Claude Code skills documentation · loads skills from .claude/skills/
Install the "api" agent skill from https://github.com/genkovich/sdd/tree/main/skills/api into .claude/skills/api/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "api", then confirm the skill loads.Claude Code copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$skill-installer install https://github.com/genkovich/sdd/tree/main/skills/apiType this inside Codex. $skill-installer <name> installs a curated skill from openai/skills. The installer writes to $CODEX_HOME/skills (default ~/.codex/skills). Restart Codex if the skill does not show up.
$ npx skills add genkovich/sdd --skill api -a codexProject install goes to .agents/skills/; add -g for ~/.codex/skills/.
$ gh skill install genkovich/sdd api --agent codexProject scope by default (.agents/skills/); add --scope user for a personal install.
$ git clone --depth 1 https://github.com/genkovich/sdd.git skills-src && mkdir -p .agents/skills && cp -r skills-src/skills/api .agents/skills/api && rm -rf skills-srcUse ~/.agents/skills/ instead of .agents/skills for a personal install.
Codex skills documentation · loads skills from .agents/skills/
Install the "api" agent skill from https://github.com/genkovich/sdd/tree/main/skills/api into .agents/skills/api/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "api", then confirm the skill loads.Codex copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$ npx skills add genkovich/sdd --skill api -a cursorProject install goes to .agents/skills/; add -g for ~/.cursor/skills/.
$ gh skill install genkovich/sdd api --agent cursorProject scope by default (.agents/skills/); add --scope user for a personal install.
$ git clone --depth 1 https://github.com/genkovich/sdd.git skills-src && mkdir -p .cursor/skills && cp -r skills-src/skills/api .cursor/skills/api && rm -rf skills-srcUse ~/.cursor/skills/ instead of .cursor/skills for a personal install.
Cursor skills documentation · loads skills from .cursor/skills/, .agents/skills/, .claude/skills/, .codex/skills/
Install the "api" agent skill from https://github.com/genkovich/sdd/tree/main/skills/api into .cursor/skills/api/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "api", then confirm the skill loads.Cursor copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$ gemini skills install https://github.com/genkovich/sdd.git --path skills/api--scope user (default) or --scope workspace; --path is the subfolder of the repo that holds the skill; --consent skips the security confirmation prompt.
$ npx skills add genkovich/sdd --skill api -a gemini-cliProject install goes to .agents/skills/; add -g for ~/.gemini/skills/.
$ gh skill install genkovich/sdd api --agent gemini-cliProject scope by default (.agents/skills/); add --scope user for a personal install.
$ git clone --depth 1 https://github.com/genkovich/sdd.git skills-src && mkdir -p .gemini/skills && cp -r skills-src/skills/api .gemini/skills/api && rm -rf skills-srcUse ~/.gemini/skills/ instead of .gemini/skills for a personal install, then run /skills reload.
Gemini CLI skills documentation · loads skills from .gemini/skills/, .agents/skills/
Install the "api" agent skill from https://github.com/genkovich/sdd/tree/main/skills/api into .gemini/skills/api/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "api", then confirm the skill loads.Gemini CLI copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$ gh skill install genkovich/sdd apiInstalls for Copilot at project scope by default; add --scope user for a personal install. Preview a skill first with gh skill preview. Needs GitHub CLI 2.90.0 or later (public preview).
$ npx skills add genkovich/sdd --skill api -a github-copilotProject install goes to .agents/skills/; add -g for ~/.copilot/skills/.
$ git clone --depth 1 https://github.com/genkovich/sdd.git skills-src && mkdir -p .github/skills && cp -r skills-src/skills/api .github/skills/api && rm -rf skills-srcUse ~/.copilot/skills/ instead of .github/skills for a personal install. Commit .github/skills so cloud agent and code review can use it.
GitHub Copilot skills documentation · loads skills from .github/skills/, .claude/skills/, .agents/skills/
Install the "api" agent skill from https://github.com/genkovich/sdd/tree/main/skills/api into .github/skills/api/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "api", then confirm the skill loads.GitHub Copilot copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$ npx skills add genkovich/sdd --skill api -a opencodeOpenCode documents no install command of its own. Project install goes to .agents/skills/; add -g for ~/.config/opencode/skills/.
$ gh skill install genkovich/sdd api --agent opencodeProject scope by default (.agents/skills/); add --scope user for a personal install.
$ git clone --depth 1 https://github.com/genkovich/sdd.git skills-src && mkdir -p .opencode/skills && cp -r skills-src/skills/api .opencode/skills/api && rm -rf skills-srcUse ~/.config/opencode/skills/ instead of .opencode/skills for a personal install.
OpenCode skills documentation · loads skills from .opencode/skills/, .claude/skills/, .agents/skills/
Install the "api" agent skill from https://github.com/genkovich/sdd/tree/main/skills/api into .opencode/skills/api/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "api", then confirm the skill loads.OpenCode copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
apiA skill your agent uses to derive the API contract for a feature — an OpenAPI 3.1 document at docs/features/{slug}/contracts/openapi.yaml plus a drift/sync report (and an events doc when the feature…
API is an agent skill from genkovich/sdd. Use to derive the API contract for a feature — an OpenAPI 3.1 document at docs/features/{slug}/contracts/openapi.yaml plus a drift/sync report (and an events doc when the feature has async flows). Triggers on "api for {slug}", "openapi for {slug}", "API contract for {slug}", "lock the interface for {slug}", "events for {slug}", "/sdd:api {slug}", "контракт API для {slug}", "OpenAPI для {slug}", "опиши ендпоінти". The contract is never hand-written: it is a derived function of data-model.md (typed fields +…
Its SKILL.md is about 4.3k tokens, which your agent loads only when the skill is triggered. The skill folder holds 5 other files, including reference files (for example `references/drift-check.md`, `templates/events.md` and `templates/openapi.yaml`).
It sits in Backend & APIs, covering OpenAPI specifications, API design and Diagrams. It works with OpenAPI. The repository describes itself as: Spec-Driven Development for Claude Code: 12 atomic Socratic skills + a TDD implement engine (agent-team & dynamic-workflow modes). The licence is MIT.
8 steps, taken from the first numbered list in SKILL.md.
Read from SKILL.md and the folder at commit 4403913. It shows what the files ask for, not the result of running them.
Pre-approves nothing: there is no allowed-tools line, so your agent's usual permission prompts apply.
From allowed-tools in the SKILL.md frontmatter.
No scripts in the folder and no shell commands in SKILL.md.
From the folder's file list and the shell code blocks in SKILL.md.
No URLs in SKILL.md.
From URLs in SKILL.md, links to its own repository left out.
Names no API keys, tokens, secrets or passwords.
From names ending in _API_KEY, _TOKEN, _SECRET, _KEY or _PASSWORD in SKILL.md.
API loads about 4.3k tokens when it runs, and up to ~6.3k if it reads all its reference files. Until then it costs about 229 tokens; SKILL.md has 1,992 words of instructions outside code blocks.
Estimates: characters ÷ 4, the usual rule of thumb; real counts depend on the model's tokenizer. Scripts and assets cost tokens only if the agent reads them.
The automated check found no risky patterns in SKILL.md.
Automated static check — not a guarantee. Review scripts before installing. It scans the text of SKILL.md for risky patterns (piping downloads into a shell, reading credential files, hidden Unicode, destructive commands); files beside SKILL.md are not scanned.
The full file from genkovich/sdd at commit 4403913, republished under its MIT licence (© genkovich). 1,992 words, ~4,258 tokens.
.claude/skills/api/SKILL.md (or your agent's skills folder). This skill also uses 3 other files; get the full folder from GitHub.Projects the upstream artifacts into one interface contract. By default that's an HTTP/OpenAPI contract; this skill is interface-kind aware — and the kind comes from the surface(s) design declared in sad.md frontmatter target_surfaces, read here, not re-derived (→ ../_shared/surfaces.md). For a non-HTTP project it produces the matching contract form (or steps aside):
contracts/openapi.yaml (OpenAPI 3.1) + api-sync-report.md..proto (or the repo's IDL) with the same derive-and-drift discipline.contracts/cli.md — the command/flag/exit-code surface derived from the AC.contracts/public-api.md — the public signatures/types the feature exposes.contracts/events.md (no request/response surface).tasks.Whatever the form, the contract is derived from data-model.md — or, on a legal no-schema-change skip, the existing schema — plus the sad.md §6 sequences + the spec's AC, never typed by hand — generation that diverges from the model or the sequences is the bug this skill exists to catch. The rest of this file details the HTTP path (the common case); the same derive → drift-check → reconcile loop applies to the other forms with the form-appropriate artifact.
This skill keeps only its own machinery. Question phrasing is shared → ../_shared/ask-style.md. Depth (events doc only when async; one resource vs full surface) follows the size matrix → ../_shared/size-matrix.md. The drift-resolution dialog reuses the shared 4-state actions — keep it short, point the machinery to _shared.
Contract summary/description prose follows artifact_language — paths, operationId, status codes and schema names never translate → ../_shared/artifact-language.md.
Backend Lead (drives the interface). The PM confirms each endpoint maps to a real user story; a frontend / consumer engineer is the first reader — the contract is locked before they start integration.
<slug> — same feature slug used by every earlier stage.docs/features/<slug>/data-model.md. When present, it is the source of typed fields and constraints. When absent, evaluate data-model's N/A condition (no schema change — size-matrix fast lane) yourself: sad.md §5 declares no new building blocks/entities, no staged docs/features/<slug>/migrations/, and the spec introduces no new entity → proceed, deriving types/constraints from the existing schema (the live migrations/ DDL + architecture-map.md §Migrations/§Conventions) and saying so loudly in the handoff. Absent and a schema change exists → STOP and point: «run data-model <slug> first — the contract is derived from its entities».sad.md frontmatter target_surfaces — picks the contract form (step 1). Absent or empty → warn («surfaces undeclared — re-run design, or proceeding as backend-service») and treat as [backend-service], falling back to the architecture-map derivation (→ ../_shared/surfaces.md).docs/features/<slug>/sad.md §6 — the Mermaid sequenceDiagram blocks. Their alt/else branches become the error responses; an async participant (<message-bus> / <external-system>) on a mutating flow marks its endpoint Idempotency-Key-required and seeds events.md. Absent → note the gap (error branches derived from spec.md §5 only — likely misses authorization branches) and still generate.docs/features/<slug>/spec.md — §4 user stories give the endpoint list; §5 acceptance criteria give the shape of each happy + error outcome. The spec deliberately holds no HTTP/status/error-code/SQL detail — that mapping is this skill's job.docs/features/<slug>/.size — depth hint. Absent → default to M (full surface) and say so loudly in the handoff — «size M (default — no .size; run /sdd:classify-size <slug>)». docs/features/<slug>/adr/*.md — override defaults (versioning, error format, auth scheme) when an ADR mandates it; CONTEXT.md (read both repo-root and docs/features/<slug>/ — per-feature wins → ../glossary/SKILL.md) — glossary terms become schema names verbatim. Existing contracts/openapi.yaml → diff and update in place, never overwrite whole-cloth.test -f docs/features/<slug>/data-model.md — a three-way gate, not a binary one:data-model's N/A condition yourself: sad.md §5 names no new building blocks/entities, no staged docs/features/<slug>/migrations/, spec introduces no new entity) → PROCEED — this is the legal fast-lane skip: derive types/constraints from the existing schema (live migrations/ DDL + architecture-map.md §Migrations/§Conventions), record each field-origin as existing schema — <migration/DDL anchor> with the same confidence scale, and state loudly in the handoff: «data-model.md absent — legal fast-lane skip (no schema change); fields derived from the existing schema»;sad.md frontmatter target_surfaces FIRST (design already declared it; the surface picks the contract form per ../_shared/surfaces.md: backend-service → OpenAPI / gRPC / events per its sub-kind; cli → contracts/cli.md; worker → contracts/events.md; library-sdk → contracts/public-api.md; a UI surface — web-frontend / mobile-app / desktop-app — consumes the backend contract, it does not author one). Fall back to deriving the kind from docs/architecture-map.md + the spec's capabilities only if the SAD or the field is absent (a greenfield run where design was skipped). HTTP/REST → the OpenAPI path below (the default, detailed here); gRPC/CLI/library/event-only → produce the matching contract form (see the intro) with this same derive→drift→reconcile loop; no external interface (pure internal logic) → skip to tasks with a one-line note in the report — this self-skip is api's N/A condition in the size-matrix fast lane. Then read data-model.md (entities, fields, types, constraints) — or, on a legal skip, the existing schema sources named above — plus sad.md §6 (flows + alt-branches + async actors), spec.md §4/§5. Surface a one-line "found / missing" note for sad.md and spec.md — never refuse on their absence, only narrow the derivation and record the gap../templates/openapi.yaml → docs/features/<slug>/contracts/openapi.yaml. If async flows exist, also ./templates/events.md → contracts/events.md. Fill info.description from spec.md §1 (why this API exists).data-model.md entity column — or, on a legal skip, to an existing-schema column (a live migration's DDL) — copy its constraints across (maxLength/pattern/enum from the model's bounded types). Never invent a field with no origin in any input — ask the user where it comes from. $ref every shared schema; no inline duplication. Lists paginate by cursor (?after=&before=&limit=), wrapped in {items, has_next, has_prev, next_cursor}.alt … else … end branch into a responses entry. The error body is the unified envelope {code, message, details?}; code follows the neutral convention module.error_name (snake_case, e.g. lesson.not_owned, lesson.invalid_state) — a naming rule, not a language artifact. Map status by class (4xx client / 5xx server). This closes the spec's usual blind spot — §5 lists the happy path + a couple of errors; the sequences enumerate the authorization and concurrent-state branches the spec omits.Idempotency-Key-required (state the TTL). For each async message, fill an events.md entry: event name module.action.vN, payload schema, producer, consumers, retry / dead-letter behaviour.<...>@example.test, +380 00 000 00 00, Test User) — never real PII.docs/features/<slug>/contracts/api-sync-report.md — see ./references/drift-check.md. It has a field-origins table (one row per operation.field: path | origin | confidence) and a checklist. The check runs both directions:spec.md §5 AC maps to ≥1 operation/response; every operation maps to a §4 user story + ≥1 AC; every sad.md §6 alt-branch has a response, and any error/authorization response the contract needs but no §6 flow shows is a sequence gap. A gap here is not an api bug — it's a hole upstream: surface it and resolve it as Save-as-OQ with the upstream stage as owner — the OQ row names the producing stage as owner (specify for a missing AC, sequences for a missing branch) with due «before the contract is finalized», so the source gets fixed through the standard 4-state machine, not a fifth action.
A core finding failing (or ≥3 flags total) pauses the run — resolve each via the shared 4-state actions (../_shared/ask-style.md): Accept / Fix (the contract) / Save-as-OQ / Drop. A fix that belongs upstream (the spec's AC, the sequence) is the Save-as-OQ variant with the upstream stage as owner (see step 7's back-feed) — never a fifth action. Never silently edit the sources — surface the mismatch and let the human pick the right artifact (the contract, the spec's AC, or the sequence).spectral lint contracts/openapi.yaml (add it to the project's check target if not yet wired). On a clean check, the files are written; propose commit api: <slug> contract. Then emit the stage-handoff block per ../_shared/handoff.md — What I did + Review (contracts/openapi.yaml, api-sync-report.md, + events.md if async) + Run next — resolve the next stage per .route (the Routes table in ../_shared/size-matrix.md): forward /sdd:screens <slug>; screens' N/A condition = no UI surface declared (sad.md target_surfaces contains none of web-frontend / mobile-app / desktop-app), skip target /sdd:tasks <slug> (on quick — auto-skip with the reason + inverted ↳ or; on standard — offer the ↳ or; on full — no skip line)./sdd:api <slug> --reconcile. Re-derives after an upstream artifact changed (typically data-model.md arrived or was tightened after a thinner first pass). It re-reads inputs, tightens loose types where the model now has a constraint, refreshes the field-origins confidence column, and — the load-bearing part — surfaces any field that had an inferred origin but now disagrees with the model. That disagreement is real drift, not stale incompleteness. info.version is never bumped silently; the user does that with a CHANGELOG line.
docs/features/<slug>/contracts/openapi.yaml written: OpenAPI 3.1, BearerAuth global with public endpoints declaring explicit security: [], every error response the {code, message, details?} envelope, every operation with examples, all shared types via $ref.api-sync-report.md written alongside: field-origins table + the 4-point drift checklist, every core finding ✓ or explicitly resolved with the user.data-model.md column (or, on a legal fast-lane skip, to an existing-schema column named in the field-origins table); every error code exists in the repo's error definitions (checked in the form the repo uses).contracts/events.md present iff the feature has async flows; each event has a payload schema, producer, consumers, retry / DLQ note.api-sync-report.md are this skill's structural self-check (../_shared/self-check.md); its result is reported in the handoff.data-model.md (keep it with a # stale note and surface it — the human decides).data-model just to produce an empty document. The gate is conditional (step 1): refuse only when a schema change actually exists; otherwise derive from the existing schema and say so.data-model.md; error codes are the neutral module.error_name convention — not a Go/TS/Python idiom and not tied to any driver's error type.{"error": "failed"}), ?v=2 query versioning, nullable: true (3.0 style — use type: [string, null]), offset pagination, or real PII in examples.design already declared it. target_surfaces in sad.md is the primary signal — read it; the architecture-map derivation is the fallback only when the SAD/field is absent (greenfield). Silently re-inferring HTTP-vs-events on every run is the double-derivation this skill's surface-awareness removes.../_shared/ask-style.md — canonical question/option phrasing for the drift-resolution dialog (step 7).../_shared/size-matrix.md — MVP (one resource, events only if async) vs Full surface depth.../_shared/surfaces.md — the declared target_surfaces (read from sad.md) pick the contract form; this skill reads, never re-derives../references/drift-check.md — the field-origins table + 4-point drift checklist, reconcile semantics, conflict table../templates/openapi.yaml — OpenAPI 3.1 scaffold: BearerAuth, cursor page wrapper, {code, message, details?} Error schema../templates/events.md — async event-contract scaffold (producer / consumers / payload / retry / DLQ).© genkovich, MIT. Rendered from Markdown: HTML in the file is shown as text, images as links, and headings moved down two levels. Raw file
SKILL.md and 3 other files (references) in skills/api of genkovich/sdd.
Open the folder on GitHubat commit 4403913
API next to the 5 skills that share the most tags, products or categories with it. Stars are the repository's; “used in” counts other GitHub owners with a copy.
| Skill | Stars | Used in | Tokens | Auto-check | Licence | Repo updated |
|---|---|---|---|---|---|---|
| API this skillgenkovich/sdd | 171 | — | ~4.3k | Automated safety check: Pass | MIT | |
| Docs Interfacesjh941213/my-cc-harness | 126 | — | ~863 | Automated safety check: Notes | None | |
| API DesignerJeffallan/claude-skills | 12k | 2 repos | ~2k | Automated safety check: Pass | MIT | |
| Old Coder API DesignAmazingAng/old-coder | 749 | 1 repos | ~3.4k | Automated safety check: Pass | MIT | |
| API Surface Reviewpolarsource/polar | 10k | — | ~1.3k | Automated safety check: Pass | MIT | |
| API ContractChenyCHENYU/Robot_Admin | 1k | — | ~1.9k | Automated safety check: Pass | MIT |
jh941213/my-cc-harness
Generate interface/API docs — OpenAPI 3.1/AsyncAPI 3.0 specs, API topology diagrams, interface flow (sequence) diagrams, API changelog.
Jeffallan/claude-skills
Designs REST and GraphQL APIs from resource modeling to an OpenAPI 3.1 contract, with versioning, pagination and RFC 7807 error handling.
AmazingAng/old-coder
Reviews or designs an HTTP/JSON API's endpoints, auth, pagination, versioning and deprecations, guarding against inventing a bespoke interface or silently breaking consumers.
polarsource/polar
Review changes to Polar's API contract — Pydantic schemas, FastAPI endpoints, OpenAPI output and the generated SDKs.
ChenyCHENYU/Robot_Admin
A skill your agent uses when: generating TypeScript API layer (type definitions + request functions) from page-spec JSON or Swagger/OpenAPI docs.
Raja0sama/vibex
Diagrams and checkable docs from a codebase. An agent skill from Raja0sama/vibex.
genkovich/sdd
A skill your agent uses to fix a reported bug spec-first: reproduce it, trace the symptom to the owning feature's acceptance criteria, pin it with a failing (RED) test, apply the minimal GREEN fix…
genkovich/sdd
A skill your agent uses to implement a feature from its tasks.json with test-driven development — writes a failing test first, makes it pass, refactors, gates, and commits per task.
genkovich/sdd
Use BEFORE roadmap or specify to get the idea OUT OF YOUR HEAD and onto disk — a Socratic interview that surfaces hidden assumptions, names tradeoffs, exposes imprecisions and proposes fresh angles…
genkovich/sdd
A skill your agent uses to classify a feature into XS/S/M/L/XL and write docs/features/{slug}/.size plus the pipeline route docs/features/{slug}/.route (quick|standard|full) so later skills know how…
genkovich/sdd
A skill your agent uses to record a post-hoc or asynchronous architecture decision as a MADR ADR when it was NOT captured during the synchronous design pass — a choice made in code, in a chat, on a…
genkovich/sdd
A skill your agent uses to break a designed feature into atomic, ≤1-day tasks with a dependency graph, a per-task Definition of Done, and a machine-readable tasks.json that the implement engine…
Works with
Categories
A skill your agent uses to derive the API contract for a feature — an OpenAPI 3.1 document at docs/features/{slug}/contracts/openapi.yaml plus a drift/sync report (and an events doc when the feature…. API is an agent skill from genkovich/sdd.yaml plus a drift/sync report (and an events doc when the feature has async flows).
API fits situations like: openapi for {slug}; API contract for {slug}; lock the interface for {slug}; events for {slug}.
Run `npx skills add genkovich/sdd --skill api -a claude-code`. Or copy the skill folder (skills/api in genkovich/sdd) into .claude/skills/api in your project. Claude Code loads it when a task matches its description.
Run `npx skills add genkovich/sdd --skill api -a codex`. Or copy the skill folder (skills/api in genkovich/sdd) into .agents/skills/api in your project. Codex loads it when a task matches its description.
Cursor, Gemini CLI, GitHub Copilot and OpenCode also load SKILL.md folders. With the skills CLI, run `npx skills add genkovich/sdd --skill api -a cursor` (or -a gemini-cli, github-copilot or opencode for the others). To copy it by hand, put the folder in .cursor/skills/api, .gemini/skills/api, .github/skills/api and .opencode/skills/api in your project.
SKILL.md names no scripts, command-line tools or credentials: API is instructions for the agent only.
SKILL.md contains no URLs. Any network use would come from the scripts or tools the agent runs. This is read from the text; nothing was executed.
Our automated static check of SKILL.md found no risky patterns, such as piping downloads into a shell, reading credential files or hidden Unicode. It is not a guarantee. Review the folder before installing.
API is published under the MIT licence (the repository's licence). It allows redistribution, so the full SKILL.md is shown on this page.
About 4.3k tokens (SKILL.md is roughly 17k characters). Agents keep only the skill's name and description in context until a task matches; then they load SKILL.md in full. Its references folder adds about 2k tokens, read only when the agent opens those files.
Skills that share tags, products or a category with API: Docs Interfaces (jh941213/my-cc-harness, 126 stars), API Designer (Jeffallan/claude-skills, 12k stars), Old Coder API Design (AmazingAng/old-coder, 749 stars) and API Surface Review (polarsource/polar, 10k stars). The comparison table on this page puts their stars, adoption, token cost, safety result and licence side by side.
genkovich (a GitHub user) maintains it in genkovich/sdd, which has 171 GitHub stars. The repository holds 21 skills in this directory. The repository was last updated on September 5, 2026.
Source: genkovich/sdd on GitHub. Facts on this page come from the repository at the commit we read; the author's words are quoted as theirs.