Project Creation Lifecycle — Gap Analysis
Date: 2026-02-18 Triggered By: Chamber of Commerce Canada (C-of-C-CANADA) project creation Author: Claude (Opus 4.6) TRACK Task: K.7.1
Executive Summary
Creating the C-of-C-CANADA venture project required 21 distinct steps across 4 sessions. The existing /research-pipeline --genesis + /new-project --from-research pipeline automates approximately 60% of these steps. The remaining 40% required manual orchestration including venture-specific submodule placement, TRACK file generation (vs TASKLIST.md), project registration in projects.db, frontmatter classification, cross-validation, and session log inception.
Recommendation: Add a --venture flag to /research-pipeline that chains a complete post-genesis lifecycle, closing the automation gap to <5%.
All 21 Steps Performed (C-of-C-CANADA)
Phase 1: Research (Steps 1-3) — AUTOMATED by /research-pipeline
| # | Step | Command/Agent | Automated? |
|---|---|---|---|
| 1 | US Market Research — 14 artifacts from 28 sources | /research-pipeline Phase 1 | YES |
| 2 | Canadian Market Extension — 2 new + 12 updated from 34 sources | Manual multi-agent orchestration | NO (no --extend-market flag) |
| 3 | Marketing/GTM/Opportunity — 3 new strategic docs + 2 updated | Manual agent dispatch | NO (Phase 3 only does ideation prompts) |
Phase 2: Visualization (Steps 4) — AUTOMATED
| # | Step | Command/Agent | Automated? |
|---|---|---|---|
| 4 | 6 JSX dashboards generated | /research-pipeline Phase 2 (with --extended) | YES |
Phase 3: Project Creation (Steps 5-12) — PARTIALLY AUTOMATED
| # | Step | Command/Agent | Automated? |
|---|---|---|---|
| 5 | Create GitHub repo (coditect-ai/chamber-of-commerce-canada) | gh repo create | PARTIAL (via /new-project Step 2) |
| 6 | Initialize local repo as venture submodule (submodules/ventures/) | Manual — /new-project uses generic category, not ventures/ | NO |
| 7 | Create .coditect + .claude symlinks | /new-project Step 2 via submodule-orchestrator | YES |
| 8 | Create CODITECT directory structure (docs/, src/, internal/, etc.) | /new-project Step 4 via project-structure-optimizer | YES |
| 9 | Copy 26 research artifacts to docs/research/ | Manual cp -r | NO |
| 10 | Create project README.md with market data from research | Manual — /new-project creates generic README | NO |
| 11 | Create PROJECT-PLAN.md from research findings | /new-project Step 3 (partial — plan from brief) | PARTIAL |
| 12 | Create .gitignore | /new-project Step 4 | YES |
Phase 4: TRACK Setup (Steps 13-14) — NOT AUTOMATED
| # | Step | Command/Agent | Automated? |
|---|---|---|---|
| 13 | Generate 5 TRACK files (R, V, D, G, O) with 107 tasks | Manual — /new-project only creates TASKLIST.md | NO |
| 14 | Generate MASTER-TRACK-INDEX.md with phase timeline | Manual | NO |
Phase 5: Registration & Compliance (Steps 15-18) — NOT AUTOMATED
| # | Step | Command/Agent | Automated? |
|---|---|---|---|
| 15 | Initial commit + push to GitHub | /new-project Step 2 (initial commit only) | PARTIAL |
| 16 | Register in projects.db (UUID, path, GitHub URL) | Manual /register-project | NO |
| 17 | Create session log at SSOT location | Manual /session-log --new-session | NO |
| 18 | Add as submodule to parent repo (coditect-rollout-master) | Manual git submodule add | NO |
Phase 6: Quality (Steps 19-21) — NOT AUTOMATED
| # | Step | Command/Agent | Automated? |
|---|---|---|---|
| 19 | Frontmatter classification (27 files, 4 parallel agents) | Manual /classify | NO |
| 20 | Cross-validation (MD5 hash of all 28 files staging vs permanent) | Manual | NO |
| 21 | Staging cleanup (remove analyze-new-artifacts/ + interim dir) | Manual (user-approved deletion) | NO |
Coverage Analysis
| Phase | Steps | Automated | Gap |
|---|---|---|---|
| Research | 3 | 1 (33%) | Market extension, strategic docs |
| Visualization | 1 | 1 (100%) | None |
| Project Creation | 8 | 4 (50%) | Venture placement, artifact copy, research-aware README, TRACK files |
| TRACK Setup | 2 | 0 (0%) | Full gap — TRACK generation doesn't exist |
| Registration | 4 | 0.5 (12%) | projects.db, session log, parent submodule |
| Quality | 3 | 0 (0%) | Full gap — classification, validation, cleanup |
| Total | 21 | ~6.5 (31%) | ~14.5 steps require manual work |
Adjusted coverage (counting partial automation): ~60% of work effort, ~31% of discrete steps.
Proposed Enhancement: --venture Flag
Design
Add --venture flag to /research-pipeline that activates a Phase 5: Post-Genesis Venture Lifecycle after the existing Phase 4 (Genesis):
/research-pipeline "topic" --venture --extended
This chains:
- Phase 0-3: Standard research pipeline (unchanged)
- Phase 4: Genesis →
/new-project --from-research --venture - Phase 5a: Venture submodule setup (GitHub repo +
submodules/ventures/placement) - Phase 5b: TRACK file generation from research artifacts (derived tracks based on project type)
- Phase 5c: Artifact placement (copy research output to
docs/research/) - Phase 5d: Research-aware README generation (executive summary + market data)
- Phase 5e: Project registration (
/register-project) - Phase 5f: Session log inception (
/session-log --new-session) - Phase 5g: Frontmatter classification (
/classify) - Phase 5h: Initial commit + push
- Phase 5i: Parent submodule registration
/new-project Changes Required
- New
--ventureflag: Places project undersubmodules/ventures/instead of generic category - New
--tracksflag: Generates TRACK-{X}-*.md files instead of TASKLIST.md - Step 3 enhancement: When
--from-research+--tracks, generate domain-specific TRACK files with tasks derived from the SDD/TDD/ADRs - Post-creation hooks: Auto-run registration, classification, session log
TRACK Generation Logic
When research artifacts include SDD + TDD + ADRs, derive TRACK files:
| Research Artifact | Derived TRACK |
|---|---|
| SDD architecture sections | D (Development/Engineering) |
| TDD API/data design | D (Development/Engineering) |
| ADRs (market strategy) | G (Go-to-Market/Sales) |
| ADRs (compliance) | O (Operations/Compliance) |
| Executive summary recommendations | V (Validation/Partnerships) |
| Research-context.json | R (Research — auto-complete) |
Suggestions for Improvement
1. Multi-Market Research Extension
Add --extend "Canada" flag to /research-pipeline that runs a secondary research phase targeting a specific market, then updates all existing artifacts with the new data. This eliminates the manual multi-agent orchestration performed in Phase 2 of C-of-C-CANADA.
2. Strategic Document Generation Phase
The current Phase 3 (Ideation) only generates follow-up prompts. Add --strategic flag or upgrade Phase 3 to also generate GTM strategy, marketing strategy, and opportunity assessment when the research topic is a market/business opportunity (vs pure technology evaluation).
3. Research-Aware README Template
Create a dedicated README template in the research-pipeline that pulls executive summary data, key market figures, and recommendation into a project-level README. Currently /new-project generates a generic README.
4. Post-Genesis Audit Trail
Automatically create a docs/provenance/ directory with:
pipeline-report.json(already exists)genesis-manifest.json(new — records all steps, timestamps, agents used)artifact-checksums.json(MD5 hashes of all generated artifacts)
5. Consolidate /project-new and /new-project
There are two overlapping commands: /project-new (simple 7-step) and /new-project (full 5-phase). Consider deprecating /project-new and adding --simple to /new-project to reduce confusion.
Files to Modify
| File | Change |
|---|---|
commands/research-pipeline.md | Add --venture flag, Phase 5 definition |
commands/new-project.md | Add --venture flag, --tracks flag, TRACK generation |
commands/register-project.md | Ensure can be invoked programmatically from pipeline |
commands/classify.md | Ensure can be invoked on directory from pipeline |
Evidence
- C-of-C-CANADA session log:
~/.coditect-data/session-logs/projects/C-of-C-CANADA/d3d3a316-09c6-8f41-4a3f-d93e422d199c/SESSION-LOG-2026-02-18.md - C-of-C-CANADA repository:
submodules/ventures/chamber-of-commerce-canada/(26 artifacts, 25,482 lines) - TRACK task: K.7.1 (this document), K.7.2-K.7.4 (implementation)
Generated by CODITECT Analysis Pipeline | AZ1.AI INC