execute

Productivity Community

Skill notes

Arguments

See .claude/skills/execute/references/options.md for complete documentation.

Required

<tasks-path>: Path to tasks directory (contains manifest.json, layer_plan.json)

Step 1: Parse Arguments

Extract all arguments from the prompt:

Step 2: Load Manifest

Read manifest.json: Extract: prd.slug: PRD identifier prd.project_path: Default project path (if not specified in args) layers: Layer definitions

Step 3: Validate Inputs

Check: Tasks path exists manifest.json exists layer_plan.json exists Project path exists (or will be created for greenfield)

Fluxly profile Runtime and permissions are inferred by Fluxly static checks, not author declarations or runtime test results.
Domain
Productivity
Author-declared agents
No explicit declaration found; this is not inferred or tested compatibility
Static check
88 / 100 · heuristic scan, not runtime safety proof
Runtime evidence
Not run; no isolated runtime test evidence is currently recorded
Author / version / license
@vinzenz · no license declared
Token usage
Heavy
Setup complexity
Guided setup
External API key
Not required
Operating systems
Unspecified (assume cross-platform)
Runtime requirements
Python
Detected file/system behavior
  • Read-only
  • Write / modify
  • Shell exec
Network behavior
External requests
Install commands
None (reference only)

Heads up: Tool permissions are not declared; actual access depends on the runtime.

01
Step 1: Parse Arguments

Extract all arguments from the prompt:

02
Step 2: Load Manifest

Read `manifest.json`: Extract: `prd.slug`: PRD identifier `prd.project_path`: Default project path (if not specified in args) `layers`: Layer definitions

03
Step 3: Validate Inputs

Check: Tasks path exists `manifest.json` exists `layer_plan.json` exists Project path exists (or will be created for greenfield)

04
Step 4: Handle State

If `--reset`: If `--resume`: If new execution:

05
Step 5: Dry Run (if requested)

If `--dry-run`: Exit after dry run output.

06
Step 6: Execute Layers

For each layer in order:

07
Step 7: Invoke Layer Agent

For each layer, call `/execute-layer`: Wait for layer completion and parse `LAYER_RESULT`.

08
Step 8: Handle Stop Condition

If any layer returns `should_stop: true`: A task was abandoned (5 failed attempts) Update state to `stopped` Output clear message with: Which task failed

09
Step 9: Report Final Status

On completion:

10
Step 10: Finalize Context (CRD Projects)

If PROJECT.md exists in the project root, update it with implemented features. Check for PROJECT.md: If exists, update context: Read all completed task XML files Extract `<exports>` sections from each task

11
Step 11: Complete

Update state: Output final summary including context update if performed:

Discussion

Powered by GitHub Discussions. Sign in with GitHub to comment, react, or subscribe.