dynamic-workflow

Create, review, run, inspect, pause, resume, and cancel durable JavaScript workflows that coordinate multiple headless Codex agents. Use for dynamic fan-out and fan-in, per-item analysis, multi-stage agent pipelines, loops or branches driven by worker results, long background runs, and ports of Claude Code dynamic workflows. Do not use for a small linear task that one Codex turn can handle directly.

Quality

89%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

Securityby

Passed

No findings from the security scan

Quality

Content

81%Weight 40%Scale 1-5

Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.

A highly actionable, well-sequenced operational guide with real validation checkpoints and feedback loops, backed by a genuine one-level reference bundle. Its weakness is repetition: authorization and callback rules appear three times over, and the inlined callback/app-server mechanics both bloat the main file and blur the progressive-disclosure boundary.

Suggestions

Move the callback/app-server mechanics (the 'Notify this Codex thread' section, CLI 0.136.0 requirements, and notify retry details) into a dedicated reference file, keeping only the arming decision tree and the launch flag rules in SKILL.md.

Deduplicate the sandbox-authorization rules that currently appear in 'Validate and obtain approval', the run section, and the 'Safety boundaries' bullet list, stating each rule once and cross-referencing it.

Collect environment prerequisites (Node.js 20+, Codex CLI 0.136.0+) into a short prerequisites note near the runner-locator block so version numbers do not pad the operational sections.

Dimension	Reasoning	Score
Conciseness	The body is dense and imperative with no explanations of known concepts, but sandbox/authorization and callback rules are restated across 'Validate and obtain approval', the run sections, and the 30+ bullet 'Safety boundaries' list, and version pins ('Codex CLI 0.136.0 or newer', 'Node.js 20 or newer') sit inline rather than in a prerequisites or deprecated section.	3 / 5
Actionability	Every phase ships copy-paste-ready material: exact bash for locating the runner, validate/run/status/resume commands with concrete flags and limits, a complete executable JavaScript workflow example with a full JSON schema, and explicit command-scoped approval instructions.	5 / 5
Workflow Clarity	The full lifecycle (locate runner → decide → author → validate → approve → run/monitor → callback handling → resume/recover) is explicitly sequenced with a validation gate before launch, a numbered six-step resume-after-edit checklist, and a post-launch decision tree that cancels and verifies when callback arming fails.	5 / 5
Progressive Disclosure	The body cleanly signals one-level-deep references to references/workflow-api.md and assets/workflow-template.js (both present in the bundle), but roughly a hundred lines of callback and app-server mechanics ('Notify this Codex thread when a detached run finishes' and its successors) are inlined in SKILL.md where a reference file would fit better.	4 / 5
	Total	17 / 20 Passed

Description

92%Weight 40%Scale 1-5

Based on the skill's description, can an agent find and select it at the right time? Clear, specific descriptions lead to better discovery.

An exemplary description: concrete verb-driven capabilities, an explicit 'Use for' trigger list covering the fan-out/pipeline/background-run patterns, and a clear negative boundary. The only gap is a handful of missing natural synonyms (e.g., 'orchestrate', 'multi-agent') in the trigger terms.

Suggestions

Add one or two common synonyms such as 'orchestrate multiple agents' or 'parallelize agent work' to the Use-for clause to widen natural trigger matching.

Dimension	Reasoning	Score
Specificity	The description enumerates seven concrete lifecycle actions ("Create, review, run, inspect, pause, resume, and cancel") on a precisely named artifact ("durable JavaScript workflows that coordinate multiple headless Codex agents"), giving comprehensive action coverage with no generic filler.	5 / 5
Completeness	It explicitly answers both what (first sentence's concrete capability list) and when ("Use for dynamic fan-out and fan-in..." plus the negative "Do not use for a small linear task that one Codex turn can handle directly") with concrete trigger phrases.	5 / 5
Trigger Term Quality	"dynamic fan-out and fan-in, per-item analysis, multi-stage agent pipelines, loops or branches driven by worker results, long background runs" are phrasings users would naturally say, but common synonyms such as "orchestrate", "parallelize", or "multi-agent coordination" are missing.	4 / 5
Distinctiveness Conflict Risk	The niche is unambiguous (durable JavaScript control flow over headless Codex workers) and the explicit do-not-use clause prevents triggering for ordinary single-turn tasks, so overlap with generic coding or orchestration skills is minimal.	5 / 5
	Total	19 / 20 Passed

Validation

100%

Warnings & errors only

Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.

Validation — 16 / 16 Passed

Validation for skill structure

No warnings or errors.

Repository: pchalasani/claude-code-tools
Path: plugins/dynamic-workflow/skills/dynamic-workflow/SKILL.md
Commit: 0ca7320

Reviewed: 14 days ago

Table of Contents

Discovery Implementation Validation

Is this your skill?

If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.