Pipeline Runner

Execute pipeline phases or custom stage ranges.

Parameters

Parse from user input:

• --phase: casual | regular | power (default: casual)
• --target: apify | smithery | npm (default: apify)
• --run: ideate | build | ship | full (default: full)
• --from: Custom starting stage (overrides --run)
• --to: Custom ending stage (overrides --run)
• --name: Spec/build name (required for build/ship phases)
• --dry-run: Skip actual publishing (default: true)

Pipeline Phases

IDEATE Phase (idea → spec)

Command: /mcp-opportunity-pipeline:run --run ideate

Runs: discover → analyze-gaps → validate → spec

Output: Validated opportunity specs ready to build

BUILD Phase (build → package)

Command: /mcp-opportunity-pipeline:run --run build --name {spec-name}

Runs: build → qa → package

Output: Tested code + docs + marketing materials

SHIP Phase (publish)

Command: /mcp-opportunity-pipeline:run --run ship --name {build-name}

Runs: publish (with dry-run by default)

Output: Live marketplace listing

FULL Pipeline

Command: /mcp-opportunity-pipeline:run --run full

Runs all 8 stages sequentially with approval gates.

Stage Reference

#	Stage	Phase	Command	Description
1	discover	IDEATE	discover	Scrape marketplaces
2	analyze-gaps	IDEATE	analyze-gaps	Score opportunities
3	validate	IDEATE	validate	Reddit pain validation
4	spec	IDEATE	spec	Generate build specs
5	build	BUILD	build	Implement MCP server
6	qa	BUILD	qa	Automated testing
7	package	BUILD	package	Generate docs/marketing
8	publish	SHIP	publish	Deploy to marketplace

Usage Examples

Ideation: Find and spec opportunities

/mcp-opportunity-pipeline:run --run ideate --phase casual

Build: Implement a specific spec

/mcp-opportunity-pipeline:run --run build --name notion-database-sync

Ship: Publish to Apify

/mcp-opportunity-pipeline:run --run ship --name notion-database-sync --dry-run false

Custom range: Just validate and spec

/mcp-opportunity-pipeline:run --from validate --to spec

Full pipeline with stops

/mcp-opportunity-pipeline:run --run full --phase casual --target apify

Process

1. Parse parameters and determine stage range

2. Check prerequisites:

   • If starting after discover, verify outputs exist
   • Load config from .claude-plugin/config.json

3. Execute stages sequentially:

   for stage in range(from_stage, to_stage + 1):
       run_stage(stage)
       check_stage_output()
       if failed:
           stop_and_report()
   

4. Report summary:

   • Stages completed
   • Outputs generated
   • Recommendations

Stage Dependencies

discover → analyze-gaps → validate → spec → build → qa → package → publish
    │           │             │        │       │      │       │
    └───────────┴─────────────┴────────┴───────┴──────┴───────┘
                      Each stage requires previous outputs

Error Handling

• Stage fails: Stop pipeline, report failure, suggest fixes
• Missing prerequisites: List missing files, suggest which stage to run
• Validation score too low: Stop at validate stage, report findings
• QA pass rate < 90%: Stop at QA, output fix suggestions

Output

Pipeline run summary saved to outputs/pipeline-run-{date}.json:

{
  "run_at": "2025-11-25T10:00:00Z",
  "phase": "casual",
  "target": "apify",
  "stages_run": ["discover", "analyze-gaps", "validate", "spec"],
  "stages_completed": ["discover", "analyze-gaps", "validate", "spec"],
  "stages_failed": [],
  "outputs": {
    "discover": "outputs/discover/raw-opportunities-2025-11-25.json",
    "analyze": "outputs/analyze/gap-opportunities-2025-11-25.json",
    "validate": "outputs/validate/validated-opportunities-2025-11-25.json",
    "spec": ["outputs/spec/notion-database-sync-spec.md"]
  },
  "recommendations": [
    "Run /mcp-pipeline:build to implement top opportunity"
  ]
}