{"id":50908626,"url":"https://github.com/maxgfr/construct","last_synced_at":"2026-06-16T08:00:31.235Z","repository":{"id":363513266,"uuid":"1263241585","full_name":"maxgfr/construct","owner":"maxgfr","description":"Turn a product idea into a grounded, buildable SRD suite — a skills.sh agent skill","archived":false,"fork":false,"pushed_at":"2026-06-09T08:04:05.000Z","size":422,"stargazers_count":0,"open_issues_count":4,"forks_count":0,"subscribers_count":0,"default_branch":"main","last_synced_at":"2026-06-09T09:12:21.404Z","etag":null,"topics":["agent-skill","ai-agents","anthropic","claude","claude-code","cli","developer-tools","greenfield","grounded-generation","prd","product-management","requirements","skills","software-requirements","srd","typescript"],"latest_commit_sha":null,"homepage":null,"language":"TypeScript","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":"mit","status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/maxgfr.png","metadata":{"files":{"readme":"README.md","changelog":"CHANGELOG.md","contributing":"CONTRIBUTING.md","funding":null,"license":"LICENSE","code_of_conduct":null,"threat_model":null,"audit":null,"citation":null,"codeowners":null,"security":null,"support":null,"governance":null,"roadmap":null,"authors":null,"dei":null,"publiccode":null,"codemeta":null,"zenodo":null,"notice":null,"maintainers":null,"copyright":null,"agents":null,"dco":null,"cla":null}},"created_at":"2026-06-08T19:02:41.000Z","updated_at":"2026-06-09T08:04:07.000Z","dependencies_parsed_at":null,"dependency_job_id":null,"html_url":"https://github.com/maxgfr/construct","commit_stats":null,"previous_names":["maxgfr/construct"],"tags_count":3,"template":false,"template_full_name":null,"purl":"pkg:github/maxgfr/construct","repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/maxgfr%2Fconstruct","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/maxgfr%2Fconstruct/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/maxgfr%2Fconstruct/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/maxgfr%2Fconstruct/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/maxgfr","download_url":"https://codeload.github.com/maxgfr/construct/tar.gz/refs/heads/main","sbom_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/maxgfr%2Fconstruct/sbom","scorecard":null,"host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":286080680,"owners_count":34396429,"icon_url":"https://github.com/github.png","version":null,"created_at":"2022-05-30T11:31:42.601Z","updated_at":"2026-05-26T15:22:16.424Z","status":"online","status_checked_at":"2026-06-16T02:00:06.860Z","response_time":126,"last_error":null,"robots_txt_status":"success","robots_txt_updated_at":"2025-07-24T06:49:26.215Z","robots_txt_url":"https://github.com/robots.txt","online":true,"can_crawl_api":true,"host_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub","repositories_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories","repository_names_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repository_names","owners_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners"}},"keywords":["agent-skill","ai-agents","anthropic","claude","claude-code","cli","developer-tools","greenfield","grounded-generation","prd","product-management","requirements","skills","software-requirements","srd","typescript"],"created_at":"2026-06-16T08:00:20.348Z","updated_at":"2026-06-16T08:00:31.230Z","avatar_url":"https://github.com/maxgfr.png","language":"TypeScript","funding_links":[],"categories":[],"sub_categories":[],"readme":"# construct\n\nTurn a product idea into a **grounded, buildable SRD suite** — a Software\nRequirements Document whose requirements and decisions rest on **real research**\n(competitors, open-source prior art, technology docs, known pitfalls), not the\nmodel's memory. A [skills.sh](https://skills.sh) agent skill.\n\n`construct` is the companion to [`reconstruct`](https://github.com/maxgfr/reconstruct)\n(rebuild a repo into PRDs) and [`ultradoc`](https://github.com/maxgfr/ultradoc)\n(answer questions grounded in a repo). Same engineering: a single committed,\n**zero-dependency** Node bundle + a thick agent playbook, fully tested, released\nby Conventional Commits.\n\n## What it does\n\n1. **Interview** the user about the product (one question at a time) → `brief.json`.\n2. **Research** the idea across keyless angles → an evidence dossier with `[E#]` ids:\n   - **market** — competitors \u0026 positioning (SearXNG → DuckDuckGo → your WebSearch),\n   - **oss** — comparable open-source projects + their issues/PRs (GitHub/GitLab),\n   - **tech** — candidate-technology docs + StackOverflow pitfalls,\n   - **semantic** *(optional)* — a local-embedding relevance pass (Qdrant + Ollama).\n3. **Analyze** the dossier: name every feature/competitor/tech/seed that would\n   render ungrounded, with the drill command that fixes each gap.\n4. **Render** a complete SRD tree (vision, scope, numbered functional requirements\n   with Given/When/Then, non-functional requirements, system context, an\n   *inferred* data model and interfaces, ADRs, competitive landscape, build plan,\n   traceability) + `SRD.json` + a machine-readable `BUILD-PLAN.json` task DAG.\n5. **Check** it: a **hard** structural-completeness gate, an **advisory**\n   grounding-coverage report (opt-in `--min-grounding` threshold), and — via\n   `review` + `check --semantic` — an opt-in **claim-support** gate that fails\n   on any cited evidence that doesn't actually back its claim.\n6. **Verify** the build *(optional)*: the agent implements the app task-by-task\n   from `BUILD-PLAN.json` (`status --json` lists the buildable task frontier so\n   independent same-milestone tasks can be built in parallel); `construct verify`\n   referees it against the SRD — declared files exist, every requirement is\n   referenced by a test, and (with `--run-tests`) the declared test commands pass.\n\nNo API keys. No `npm install` at skill-use time.\n\n## Install\n\n```\nnpx skills add maxgfr/construct\n```\n\n## Use (standalone CLI)\n\n```\nnode scripts/construct.mjs init --idea \"a self-hosted read-it-later app\" --out ./readpile\n# …fill ./readpile/brief.json via the interview…\nnode scripts/construct.mjs research --out ./readpile --angles market,oss,tech\nnode scripts/construct.mjs analyze  --out ./readpile          # what's thin? drill it\nnode scripts/construct.mjs render   --out ./readpile --level complex\nnode scripts/construct.mjs check    --out ./readpile          # add --min-grounding 70 to enforce\nnode scripts/construct.mjs review   --out ./readpile          # adjudicate each cited [E#] → verdicts.json\nnode scripts/construct.mjs check    --out ./readpile --semantic   # gate refuted/unsupported claims\nnode scripts/construct.mjs verify   --out ./readpile --app ./readpile-app --run-tests --strict\n```\n\nAdd `--merge` to also emit a single-file `SRD.md`. Run `--help` for the full\nsurface, or see [`SKILL.md`](SKILL.md) for the agent playbook and\n[`DOCUMENTATION.md`](DOCUMENTATION.md) for internals.\n\n## Output\n\n```\n00-overview/   VISION.md · SCOPE.md\nrequirements/  FUNCTIONAL.md · NON-FUNCTIONAL.md\narchitecture/  SYSTEM-CONTEXT.md · DATA-MODEL.md · INTERFACES.md · decisions/NNNN-*.md\ncompetitive/   LANDSCAPE.md\nBUILD-PLAN.md · BUILD-PLAN.json (task DAG for the build phase) · TRACEABILITY.md\nevidence/      EVIDENCE.md · evidence.json · meta.json   ·   brief.json · SRD.json\nVERIFY.md · VERIFY.todo.json · VERIFY.json (claim-support review, from `review`)\n```\n\n## Grounding is advisory; completeness is enforced\n\n`construct check` separates the two axes. The **structural gate** fails the build\non an incomplete SRD (unresolved `🧠` decisions, an FR with no acceptance\ncriteria, a dangling reference, a missing required NFR category, a malformed\nADR). The **grounding coverage** is a report — it tells you how well-cited the\nSRD is so you can invest research where it matters, but it never fails the build\nby default. When you *do* want it enforced, `--min-grounding \u003c0-100\u003e` opts into\na second gate that fails below the threshold.\n\nCoverage counts citations; it does not check they *hold*. `construct review`\nbuilds a claim↔evidence worklist (one pair per cited `[E#]`); an agent (or a\nfan-out of skeptic subagents — see `references/orchestration.md`) adjudicates\neach as `supported | partial | refuted | unsupported`, and `check --semantic`\nturns that into a third opt-in gate that fails on a refuted or unsupported\nclaim.\n\n## Optional local stack\n\n```\nnode scripts/construct.mjs semantic up   # Qdrant + Ollama + SearXNG, fully local, no key\n```\n\nSee [`references/semantic-setup.md`](references/semantic-setup.md).\n\n## License\n\nMIT © maxgfr\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fmaxgfr%2Fconstruct","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fmaxgfr%2Fconstruct","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fmaxgfr%2Fconstruct/lists"}