CtrlK
BlogDocsLog inGet started
Tessl Logo

dmux-workflows

Multi-agent orchestration using dmux (tmux pane manager for AI agents). Patterns for parallel agent workflows across Claude Code, Codex, OpenCode, and other harnesses. Use when running multiple agent sessions in parallel or coordinating multi-agent development workflows.

62

Quality

72%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

SecuritybySnyk

Passed

No findings from the security scan

Fix and improve this skill with Tessl

tessl review fix ./.agents/skills/dmux-workflows/SKILL.md
SKILL.md
Quality
Evals
Security

Quality

Content

75%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 well-structured, lean skill body with executable commands, concrete per-pane prompt examples, and useful operational guidance (worktrees, token limits, troubleshooting). The main gaps are mechanical: merge/integration steps are described as UI keypresses rather than commands, and there is no explicit post-merge validation step.

Suggestions

Show how merging works mechanically or how to feed Pane 1's findings file into Pane 2 (e.g., 'tell Pane 2: read /tmp/rate-limit-research.md'), since 'merge findings into Pane 2's context' is currently unexplained.

Add an explicit post-merge validation checkpoint (e.g., 'after merging, run the full test suite in the main pane before committing') to close the workflow loop.

Consider moving the five workflow patterns into a references/patterns.md file to keep SKILL.md as a tighter overview.

DimensionReasoningScore

Conciseness

The body is lean throughout: 'Press n to create a new pane', 'Press m to merge pane output back', the patterns are short prompt blocks, and Best Practices are one-line imperatives. It doesn't explain concepts Claude already knows — even the 'What is dmux' section only states tool facts Claude wouldn't know. Not a 5 because 'What is dmux' partially overlaps 'Quick Start' and the Complementary Tools table adds peripheral content that could be trimmed; not a 3 because there is no noticeable padding or over-explanation.

4 / 5

Actionability

Concrete, runnable commands are present: 'npm install -g dmux', 'dmux', 'git worktree add ../feature-auth feat/auth', 'git merge feat/auth', 'brew install tmux', plus copy-paste-ready agent prompts per pane. Not a 5 because key mechanics are described only at UI level ('press m to merge') with no scripted equivalent, and Pattern 1's 'merge findings into Pane 2's context' gives no instruction on how; not a 3 because most guidance is directly executable with specific examples covering the common cases.

4 / 5

Workflow Clarity

The Quick Start sequences the core loop (start dmux -> create panes with prompts -> press m to merge), each pattern includes a follow-up step, and checkpoints appear in Best Practices ('Merge strategically. Review pane output before merging to avoid conflicts') and Troubleshooting (merge conflicts -> use worktrees; unresponsive pane -> use m to read output). Not a 5 because post-merge integration has no explicit validate-after-merge step or feedback loop; not a 3 because the sequence is clear and pre-merge review plus error-recovery guidance are explicitly stated.

4 / 5

Progressive Disclosure

The skill is a single well-organized file with clear section headers (When to Activate, Quick Start, Workflow Patterns, Best Practices, Git Worktree Integration, Troubleshooting) and no nested references — one level deep by construction. Not a 5 because at ~140 lines the five workflow patterns and troubleshooting content are candidates for a references/ file that would keep SKILL.md as a tighter overview; not a 3 because nothing that clearly belongs in a separate file is inlined and navigation via headers is easy.

4 / 5

Total

16

/

20

Passed

Description

70%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.

A solid description: it names a specific tool, states the domain, and includes an explicit 'Use when' clause with natural trigger terms. Its main weakness is that capabilities stay at the 'orchestration/patterns' abstraction level rather than naming concrete actions, and the trigger conditions could include more concrete quoteable phrases.

Suggestions

Enumerate 2-3 concrete capabilities in the description (e.g., 'create agent panes, run parallel sessions across Claude Code/Codex/OpenCode, merge results back') instead of the abstract 'Multi-agent orchestration' framing.

Add concrete trigger phrases to the 'Use when' clause, e.g., 'Use when the user says "run in parallel", "split this work", or asks to coordinate multiple agent sessions'.

Include a couple of common synonyms such as 'subagents' or 'parallel agents' to widen natural trigger coverage without adding bulk.

DimensionReasoningScore

Specificity

Quotes: "Multi-agent orchestration using dmux (tmux pane manager for AI agents)" and "Patterns for parallel agent workflows across Claude Code, Codex, OpenCode". It names the domain and the dmux tool with 1-2 concrete capabilities (pane management, parallel agent workflows), but doesn't enumerate specific actions like creating panes, merging output, or worktree isolation — not comprehensive. Not a 4 because the actions described are still at the 'orchestration/patterns' level rather than several specific concrete actions; not a 2 because it does name concrete domain actions beyond a bare domain label.

3 / 5

Completeness

Quotes: what — "Multi-agent orchestration using dmux (tmux pane manager for AI agents). Patterns for parallel agent workflows..."; when — "Use when running multiple agent sessions in parallel or coordinating multi-agent development workflows." Both what and when are explicitly present. Not a 5 because the 'when' clause lacks concrete quoteable trigger phrases (e.g., 'user says run in parallel' or 'split this task'); not a 3 because the 'when' is explicit, not merely implied.

4 / 5

Trigger Term Quality

Quotes: "multi-agent", "parallel agent workflows", "agent sessions in parallel", "Claude Code", "Codex", "OpenCode", "tmux", "dmux". Good coverage of natural terms a user would say when wanting parallel agents, including harness names. Not a 5 because common synonyms like "subagents", "split this work", or "fan out" are absent; not a 3 because multiple natural trigger phrases and tool names are present.

4 / 5

Distinctiveness Conflict Risk

Quotes: "using dmux (tmux pane manager for AI agents)" — a clearly named niche tool with distinct triggers. Mostly distinct from generic orchestration skills; minor overlap risk with built-in multi-agent/subagent tooling skills, since 'multi-agent orchestration' and 'coordinating multi-agent development workflows' are broad phrases that could also match those. Not a 5 due to that overlap risk; not a 3 because the dmux/tmux-pane framing and harness names anchor it to a specific tool.

4 / 5

Total

15

/

20

Passed

Validation

100%

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
affaan-m/ECC
Reviewed

Table of Contents

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.