CtrlK
BlogDocsLog inGet started
Tessl Logo
SkillPublic

coding-cli-management

Execute AI coding CLI tools (Claude Code / Gemini CLI / qodercli / Qwen Code / OpenCode) on behalf of Workers. Use when a Worker sends a coding-request: message, asking the delegating agent (Manager, team Leader, or an authorized Worker) to run coding operations in their workspace.

Quality

80%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide
SecuritybySnyk

Critical

Do not install without reviewing

Fix and improve this skill with Tessl

tessl review fix ./manager/agent/skills/coding-cli-management/SKILL.md
SKILL.md
Quality
Evals
Security

Quality

Content

63%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.

The body is highly actionable with a well-validated delegation workflow, but token efficiency is poor: massive repetition and inline version-drift field notes inflate it to 470 lines where the reference file exists precisely to hold that detail. Moving the version/install/network forensics into references/delegation-guidance.md would fix both conciseness and progressive disclosure at once.

Suggestions

Deduplicate the --json-file drift material: state the capture-by-default contract once (at the flag table) and remove the near-identical repetitions in the Runner notes paragraph and Model configuration section.

Move the version-channel/install-matrix/Node-OpenSSL/PAGESIZE field notes (roughly lines 71-106) into references/delegation-guidance.md, keeping only a one-line pointer plus the currently-relevant version in SKILL.md.

Collect time-sensitive version numbers and field-verification dates (0.25.0/0.25.1/0.26.3, 2026-10-30, etc.) into a single versioned 'field-verified reference' table in the reference file so the main body does not go stale.

DimensionReasoningScore

Conciseness

The 470-line body is noticeably verbose: the --json-file drift material is repeated nearly verbatim in three places (the Runner notes paragraph, the flag bullet, and Model configuration), and ~130 lines of version-specific field notes ("field-verified 2026-10-30", "0.24.7-0.26.3", OpenSSL/Raspberry-Pi PAGESIZE forensics) carry inline dates and version numbers the rubric explicitly penalizes outside a deprecated/old-patterns section. It avoids a 1 because the material is domain-specific operational content, not concepts Claude already knows.

2 / 5

Actionability

Fully executable throughout: complete copy-paste bash procedures (workspace sync, marker coordination, the run-coding-cli.sh invocation with all flags), concrete config JSON schemas, the management API endpoint, and ready-to-use message and spec.md templates covering the common cases.

5 / 5

Workflow Clarity

Steps 1-4 are clearly sequenced with explicit validation (processing-marker check, exit-code branch, success-only workspace push) and a real error-recovery loop (Step 4 failure notification with a diagnostic checklist for admin and Worker). It falls short of 5 because 100+ lines of version-drift digressions sit between the workflow steps and Step 2 is thin relative to the rest.

4 / 5

Progressive Disclosure

The bundle structure is sound — references/delegation-guidance.md exists, matches its advertised Part A/Part B description, is one level deep and clearly signaled, and the three scripts are referenced by path. However, a large fraction of the body (version channels, install matrix, Node/OpenSSL generation notes, JSON-file drift measurements) is exactly the operational detail that belongs in the reference file, which goes beyond the "minor organization gaps" of the next anchor up.

3 / 5

Total

14

/

20

Passed

Description

87%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 strong description: third-person, concise, with an explicit what-and-when structure anchored by a concrete protocol trigger and enumerated tool names. The only weakness is that it describes essentially one capability (delegated CLI execution) rather than a comprehensive set of actions.

DimensionReasoningScore

Specificity

"Execute AI coding CLI tools (Claude Code / Gemini CLI / qodercli / Qwen Code / OpenCode) on behalf of Workers" names a concrete action with enumerated tools and explicit roles, but only covers 1-2 distinct actions rather than the comprehensive multi-action coverage of the top anchor.

4 / 5

Completeness

Explicitly answers both what ("Execute AI coding CLI tools ... on behalf of Workers") and when ("Use when a Worker sends a coding-request: message") with a concrete trigger phrase, matching the top anchor pattern.

5 / 5

Trigger Term Quality

Includes the concrete protocol trigger "coding-request: message" plus all five CLI names a user would naturally say, but misses common synonyms such as "delegate coding" or generic phrases like "run the CLI for me".

4 / 5

Distinctiveness Conflict Risk

The "on behalf of Workers" scope plus the distinctive "coding-request:" message trigger carve a clear niche that is unlikely to fire for the Worker-owned coding-cli skill or unrelated skills.

5 / 5

Total

18

/

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

agentscope-ai/AgentTeams

Path

manager/agent/skills/coding-cli-management/SKILL.md

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.