Index of every page on this site: llms.txt
All skills
Run it yourself

Turn any project (a requirements doc or a one-paragraph idea) into a canonical, Claude-Code-buildable, gated Basecamp build plan - a single story-driven to-do list with a 🤖 AI / 🧑 Human task split, an assigned builder, milestone approval gates for an approver, and 5 linked Docs & Files. Invoke when Ali says "story-build this", "run the story-driven build", "turn this doc into a project", hands over a build requirements document, or wants a project made assignable to an intern in Basecamp. This is the same process that produced Detroit, VA ERP, Selective Service, Fairfax, Social Pilot AI, and IPOS.

  • 8 files
  • 74.2 KB
  • Updated 3 days ago
  • GitHub
Use this Skill: https://skilld.dev/gh/colaberryintern/colaberryenterprise_ai_leadershipaccelerator/story-build

Nothing lands on disk. Nothing to clean up.

Fork this Skill

Edit a local copy. It keeps the author and licence.

MYDAY_SYNC.md

≈1.2k tokens on demand. Your agent reads this file only when SKILL.md points to it.

Prompt: make "Create a project on My Day" follow the full Story-Driven Build

Paste the block below into a Claude Code session running in the AI Project Architect & Build Companion repo (the advisor / FastAPI repo that owns execution/advisory/deep_plan.py). It is discovery-first: it tells that session what "fully created" means and has it wire the create-project path to reach that state, idempotently, instead of stopping halfway.


You are working in the AI Project Architect repo. The Basecamp side of the Story-Driven Build already works: for recent projects a story-driven to-do list and its tasks are already created in Basecamp. The gap is that creating a project on My Day does not run the whole process to completion - it must, and it must be idempotent.

What "fully created" means (definition of done)

A project is only fully created when ALL of these are true and verifiable:

  1. A gated plan exists. The 6-stage pipeline (execution/advisory/deep_plan.py) has produced the plan and it passes all four gates: (a) zero no-code mentions, (b) >= 12 stories AND >= 3 releases, (c) trace.ok == true, (d) strong real-code signal (React/Node/PostgreSQL/Docker). Claude Code / real code is assumed; no no-code tools.
  2. Basecamp reflects it. One story-driven to-do list with per-release groups; every story is a to-do carrying Fulfills/Owner-agent/Gherkin/Build/Trust/Loop-stop, marked 🤖 [AI] or 🧑 [Human], assigned to the builder; a MILESTONE APPROVALS group assigned to the approver (per-release or phase-gate mode); and 5 Docs & Files documents (Requirements, Architecture, Trust/TBI Primer, Build Guide, Traceability Matrix) deep-linked from every task.
  3. My Day reflects it. The project and its tasks appear as native My Day items, the 🤖/🧑 markers drive the My Day tier split (task_kind), the builder sees their build tasks and the approver sees their milestone gates, and the 5 documents are reachable.
  4. The state is consistent. No task exists in Basecamp that is missing from My Day (or vice-versa); re-running the create/sync produces the same end state with no duplicates.

Your task

  1. Map the current flow. Find the "Create a new project" / My Day project-creation entry point and everything it calls: the deep_plan generator, the publisher (deep_plan_publisher.py or the pluggable-publisher path), and the sync that pushes into My Day / the portal (the build-sync webhook / adapter). Write a short map: entry point -> generate -> gate -> publish (Basecamp) -> sync (My Day) -> mark done. Note exactly where the chain currently stops short of the definition of done above.

  2. Make create-project run the whole chain to completion. The create flow should drive: generate -> enforce the 4 gates (refuse/repair on fail, do not ship a half-plan) -> publish to Basecamp with the AI/Human split, builder assignment, approver gates, and 5 linked docs -> sync into My Day -> only then report the project as created. It must not mark the project created until a completion check passes.

  3. Add an explicit completion check + reconcile. Implement a verify_fully_created (or extend the existing status) that checks all four conditions above against the live Basecamp list and the My Day project, returns a per-condition pass/fail, and a reconcile step that fills only the missing pieces (missing tasks, missing 🤖/🧑 markers, missing docs, unsynced My Day items). Bound the repair loop; if it cannot converge, surface a clear escalation with the failing conditions rather than looping.

  4. Idempotency is non-negotiable. Creating/syncing the same project twice must produce the same end state with no duplicate lists, tasks, docs, or My Day items. Key each side effect (Basecamp task, My Day task, doc) so re-runs upsert, never append. Never write to Basecamp as anyone but Ali (guard on /my/profile.json id 17454835, or the service identity the sync is supposed to use); halt if the identity is wrong.

  5. Verify and report. Add/extend tests that assert the completion check catches each missing-piece case, and run the create flow against one real project end to end. Report the before/after map, the files changed, and a completion-check transcript proving all four conditions pass. Follow this repo's idempotency, failure-first, and contract rules.

Reference implementation to mirror (already proven on Detroit, Social Pilot, IPOS): the Colaberry Accelerator skill .claude/skills/story-build/scripts/publish_story_build.js shows the exact Basecamp structure, the AI/Human split rule, the builder + approver-gate model, the 5-doc set, the 4 gates, and the identity guard. Match that structure on the My Day side.

Source: SKILL.md on GitHub

skilld matched fixed text patterns in SKILL.md and file names. Patterns miss obfuscated code.

skilld run checks every file with the same patterns. It asks for approval before it loads a Skill with a behavior marked Needs approval.

No third-party reports yet.

Signed by skilld at 7f168b8. This ties the file your Agent reads to that commit on GitHub. It does not review the instructions.

Last checked against GitHub yesterday.

Activeupdated 3 days ago

README badge

README badge for colaberryintern/colaberryenterprise_ai_leadershipaccelerator/story-build