agentleFS
fortunto2fortunto2/solo-factory/skills/pipeline/SKILL.md

solo-pipeline

Skill18 starsChanged 4 months agoMITView on GitHub

Use when "run pipeline", "automate research to PRD", "full pipeline", "research and validate", "scaffold to build", "loop until done", "chain skills", or need multi-skill automation. Do NOT use for single skills (use the skill directly).

The file

We split this skill into the rules it gives an agent

8 rules. About 0 are common, 5 are distinctive, and 3 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈5

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 3

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    If no arguments or unclear, ask:

    line 41
  2. Distinctive:

    Run the first skill in the pipeline directly:

    line 66
  3. Very specific:

    If you have the solo-factory plugin installed, launcher scripts provide tmux dashboard and logging:

    line 76
  4. Distinctive:

    When all stages are done, output:

    line 88
  5. Very specific:

    When launched from terminal (without --no-dashboard), a tmux dashboard opens automatically with:

    line 109
  6. Distinctive:

    No need to close/recreate — just run the same command again

    line 139
  7. Distinctive:

    Always confirm: before starting a pipeline.

    line 157
  8. Very specific:

    Use --no-dashboard: when running from within Claude Code skill context.

    line 161

≈ “Common” is an estimate: at least 1 in 5 of the 372,429 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

/pipeline

Launch an automated multi-skill pipeline. The Stop hook chains skills automatically — no manual invocation needed between stages.

Available Pipelines

Research Pipeline

/pipeline research "AI therapist app"

Chains: /research -> /validate Produces: research.md -> prd.md

Dev Pipeline

/pipeline dev "project-name" "stack" /pipeline dev "project-name" "stack" --feature "user onboarding"

Chains: /scaffold -> /setup -> /plan -> /build Produces: full project with workflow, plan, and implementation

Steps

1. Parse Arguments

Extract from $ARGUMENTS:

  • Pipeline type: first word (research or dev)
  • Remaining args: passed to the launcher script

If no arguments or unclear, ask:

Which pipeline do you want to run?

1. Research Pipeline — /research → /validate (idea to PRD)
2. Dev Pipeline — /scaffold → /setup → /plan → /build (PRD to running code)

2. Confirm with User

Show what will happen:

Pipeline: {type}
Stages: {stage1} → {stage2} → ...
Idea/Project: {name}

This will run multiple skills automatically. Continue?

Ask via AskUserQuestion.

3. Start First Stage

Run the first skill in the pipeline directly:

For research pipeline: Run /research "idea name" For dev pipeline: Run /scaffold project-name stack

The Stop hook (if configured) will handle subsequent stages automatically. Without a Stop hook, manually invoke each skill in sequence.

3b. Launcher Scripts (optional, Claude Code plugin only)

If you have the solo-factory plugin installed, launcher scripts provide tmux dashboard and logging:

# Only available with Claude Code plugin — skip if not installed
solo-research.sh "idea name" [--project name]
solo-dev.sh "project-name" "stack" [--feature "desc"]

Pass --no-dashboard when running from within a skill context.

5. Pipeline Completion

When all stages are done, output:

<solo:done/>

The Stop hook checks for this signal and cleans up the state file.

State File

Location: .solo/pipelines/solo-pipeline-{project}.local.md (project-local) or ~/.solo/pipelines/solo-pipeline-{project}.local.md (global fallback) Log file: .solo/pipelines/solo-pipeline-{project}.log

Format: YAML frontmatter with stages list, project_root, and log_file fields. The Stop hook reads this file on every session exit attempt.

To cancel a pipeline manually: delete the state file solo-pipeline-{project}.local.md

Monitoring

tmux Dashboard (terminal use)

When launched from terminal (without --no-dashboard), a tmux dashboard opens automatically with:

  • Pane 0: work area
  • Pane 1: tail -f on log file
  • Pane 2: live status display (refreshes every 2s)

Manual Monitoring

Monitor pipeline progress with standard tools:

# Watch log file
tail -f .solo/pipelines/solo-pipeline-<project>.log

# Check pipeline state

# Auto-refresh
watch -n2 -c solo-pipeline-status.sh

Otherwise, use standard tools:

# Log tail
tail -f .solo/pipelines/solo-pipeline-<project>.log

# Check state file
cat .solo/pipelines/solo-pipeline-<project>.local.md

Session Reuse

Re-running a pipeline reuses any existing state — completed stages are skipped automatically.

  • No need to close/recreate — just run the same command again

Log Format

[22:30:15] START    | my-app | stages: research -> validate | max: 5
[22:30:16] STAGE    | iter 1/5 | stage 1/2: research
[22:30:16] INVOKE   | /research "AI therapist app"
[22:35:42] CHECK    | research | .../research.md -> FOUND
[22:35:42] STAGE    | iter 2/5 | stage 2/2: validate
[22:35:42] INVOKE   | /validate "AI therapist app"
[22:40:10] CHECK    | validate | .../prd.md -> FOUND
[22:40:10] DONE     | All stages complete! Promise detected.
[22:40:10] FINISH   | Duration: 10m

Critical Rules

  1. Always confirm before starting a pipeline.
  2. Don't skip stages — the hook handles progression.
  3. Cancel = delete state file — tell users this if they want to stop.
  4. Max iterations prevent infinite loops (default 5 for research, 15 for dev).
  5. Use --no-dashboard when running from within Claude Code skill context.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.