ask

Send a request to a CCB agent with `ask`.

Quality

71%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

Securityby

High

Do not use without reviewing

Fix and improve this skill with Tessl

tessl review fix ./inherit_skills/claude_skills/ask/SKILL.md

Quality

Content

88%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 lean, highly actionable instruction skill: an explicit trigger, a decision checklist, concrete command forms with heredocs, error-recovery paths, and a defined end-state. Its only real weakness is the long flat Guardrails list, which repeats a couple of `--chain` rules and would benefit from grouping.

Dimension	Reasoning	Score
Conciseness	The body is dense operational guidance ("Default: do not use `--chain`", "Finish an inbound CCB task in its current turn") with no padding explaining concepts Claude already knows. Minor trimming is possible — the `--chain` dependency rule is restated in several guardrail bullets ("Never add `--chain` merely to make a rejected plain ask succeed", "Use `--chain` only for a real child dependency") — which keeps it below a 5.	4 / 5
Actionability	Copy-paste-ready command templates ("command ask --chain --artifact-reply \"$TARGET\" <<'EOF'"), exact flag semantics, a quoted error string to react to ("If CCB says `ask --chain requires an active parent job`, retry once with plain `ask`"), and a concrete correction command (`ccb followup <active_job_id> --message ...`) with success/failure states. It fully specifies what to execute and what to do on each outcome.	5 / 5
Workflow Clarity	A numbered Decision Card checklist (need delegation → dependency gate → result intent → request fidelity) sequences the decision, guardrails cover error recovery (retry with plain `ask`, `injected` vs `rejected`/`too_late`/`terminal` handling), and the post-command instruction ("end the turn. Do not wait for a reply ... do not poll") closes the loop. Checklists plus explicit feedback loops match the top anchor.	5 / 5
Progressive Disclosure	No bundle files exist, and the body is organized into clear sections (trigger, Decision Card, Guardrails, templates, post-command behavior) with no nested references. The ~20-bullet Guardrails wall is one block that could be grouped into sub-sections (chain rules, artifact rules, diagnostics) for easier scanning, which keeps it below a 5.	4 / 5
	Total	18 / 20 Passed

Description

53%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.

The description is concise and correctly third-person, naming a clear niche, but it is minimal: one action, no usage triggers, and no synonyms. It would rarely trigger on its own; the body's 'Use this when the user asks you to delegate with CCB' clause should be surfaced into the description.

Suggestions

Add an explicit trigger clause to the description, e.g. 'Use when the user asks to delegate, hand off, or send work to another agent with CCB, or when project memory says to use CCB ask.'

Mention 1-2 more concrete capabilities (e.g. routing modes like `--chain`, `--silence`, `--compact`, or artifact-based request/reply) so coverage goes beyond the single 'send a request' action.

Include natural synonyms users would say (delegate, dispatch, message an agent) alongside 'send a request' to improve trigger matching.

Dimension	Reasoning	Score
Specificity	"Send a request to a CCB agent with `ask`" names the domain (CCB delegation) and one concrete action (send a request via `ask`). It stops at a single action — no mention of the routing modes the skill actually governs (`--chain`, `--silence`, `--compact`, artifact flags) — matching 'names domain and 1-2 concrete actions, but not comprehensive', not the level-4 'several specific actions'.	3 / 5
Completeness	The 'what' is clear (send a request to a CCB agent using `ask`) but the description contains no 'Use when...' clause or equivalent trigger guidance — that guidance lives only in the body ("Use this when the user asks you to delegate with CCB"). Per the judging guidelines, a missing explicit 'when' caps completeness at 3.	3 / 5
Trigger Term Quality	Relevant terms are present ("request", "CCB agent", "ask"), but common variations a user might naturally say — "delegate", "message", "hand off", "collaborate" — are absent. This is 'some relevant keywords but missing common variations or synonyms' rather than the good-coverage level 4.	3 / 5
Distinctiveness Conflict Risk	"CCB agent" and the `ask` command name a specific niche unlikely to fire for unrelated skills. It is not a 5 because the bare phrase "send a request to an agent" is generic and could overlap with other delegation/messaging skills if the CCB qualifier isn't in the user's phrasing.	4 / 5
	Total	13 / 20 Passed

Validation

93%

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 — 15 / 16 Passed

Validation for skill structure

Criteria	Description	Result
metadata_version	'metadata.version' is missing	Warning

	Total	15 / 16 Passed

Repository: SeemSeam/claude_codex_bridge
Path: inherit_skills/claude_skills/ask/SKILL.md
Commit: 18bf0af

Reviewed: 7 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.