Skip to content
Back to skills

Execution Engine

ASecurity

Run a multi-source CocoFlow research pipeline. Supports `$flow research` with linear or deep mode.

  • 724 stars
  • 0 votes
  • 0 copies
  • 0 views
  • Added September 5, 2026
ai-agents

Security analysis

A100/100

Pro scans all 20 files and shows the line behind each finding

Scanned September 5, 2026

npx -y skills add Snowflake-Labs/cocoplus --skill execution-engine --agent claude-code

Installs into .claude/skills of the current project.

Are you the author of Execution Engine?

Add the live security badge to your README. It updates with every re-scan.

Security grade badge for Execution Engine
[![Security: A — Skills Directory](https://www.skillsdirectory.com/api/skills/snowflake-labs-execution-engine-30a831af/badge)](https://www.skillsdirectory.com/skills/snowflake-labs-execution-engine-30a831af)

More formats (shields.io, HTML) on the badges page. Keep it an A: scan every change in CI with Pro.

Download with Pro
SKILL.md
---
name: "flow-research"
description: "Run a multi-source CocoFlow research pipeline. Supports `$flow research` with linear or deep mode."
version: "2.0.0"
author: "CocoPlus"
tags:
  - cocoplus
  - cocoflow
  - research
---

Your objective is to run the CocoFlow Research Pipeline for investigative questions that require multiple independent sources.

## Pipeline

1. Planner decomposes the question into source-specific subquestions.
2. Human approves the retrieval plan before crawlers launch.
3. Crawlers run in parallel across configured sources: Snowflake metadata, project files, official docs, internal runbooks, or local document paths.
4. Aggregator combines source-tracked findings.
5. Publisher writes the report with citations.
6. Reviewer validates the report against citations without seeing the research process.
7. Revisor updates the report until Reviewer is satisfied.

## Modes

- Linear mode: one Planner -> Crawler group -> Aggregator -> Publisher pass.
- Deep mode: each crawler result may produce follow-up subquestions. Depth defaults to 1 and may be configured up to 3. Breadth defaults to 2 and may be configured up to 5.

## Configuration

Read `[research]` and `[session]` from `cocoplus.toml` when present:

- `local_doc_paths`
- `default_mode`
- `max_depth`
- `max_breadth`
- `require_plan_approval`

## Exit Criteria

- [ ] The approved plan is recorded before retrieval.
- [ ] Every claim in the report traces to at least one source.
- [ ] Reviewer and Revisor are separated from the crawler process.

## Anti-Rationalization

| Shortcut / Temptation | Why It Fails |
|-----------------------|--------------|
| Treat the skill as complete because the file exists | Skill contracts must describe observable behavior and verification, not just command names. |
| Skip artifact and safety checks for a small command | Small commands still mutate state or guide execution; preserve the same gates. |

Files in this skill

  • flow-event-reader.skill.md3.1 KB
  • flow-gate-clear.skill.md1.6 KB
  • flow-gate-status.skill.md1.4 KB
  • flow-pause.skill.md1.2 KB
  • flow-research.skill.md1.9 KB
  • flow-resume.skill.md1.6 KB
  • flow-run.skill.md23.3 KB
  • flow-settle.skill.md2.1 KB
  • flow-status.skill.md3.4 KB
  • flow-template.skill.md3.4 KB
  • flow-view.skill.md4.3 KB
  • model-tier-resolve.skill.md1.6 KB
  • noop-check.skill.md1.5 KB
  • runtime-queue.skill.md3.3 KB
  • templates/drain-loop.md392 B
  • templates/hitl-gate.md358 B
  • templates/litmus-test.md398 B
  • templates/noop-workflow.md441 B
  • templates/per-project-working-directory.md408 B

Attribution

Is this your skill, or is something wrong with this listing? Request removal or report an issue. Author removals are honored within 72 hours.

Comments

Loading comments…