CtrlK
BlogDocsLog inGet started
Tessl Logo
SkillPublic

agentflow

Orchestrate autonomous AI development pipelines through your Kanban board (Asana, GitHub Projects, Linear). Manages multi-worker Claude Code dispatch, deterministic quality gates, adversarial review, per-task cost tracking, and recoverable pipeline design.

Quality

72%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide
SecuritybySnyk

Medium

Suggest reviewing before use

Fix and improve this skill with Tessl

tessl review fix ./skills/antigravity-agentflow/SKILL.md
SKILL.md
Quality
Evals
Security

Quality

Content

67%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 well structured and actionable, with concrete commands, explicit stage gates, and sensible delegation of deep detail to upstream docs. Its weaknesses are repeated safety caveats that pad the token cost and a setup workflow that could use explicit verification checkpoints.

Suggestions

Consolidate the lease/heartbeat/untrusted-data caveats, which currently appear in Core Concepts, Prerequisites, and Troubleshooting, into a single Safety section to reduce token cost.

Add a verification step between setup and dispatch (e.g., confirm the board integration and one worker lease before starting parallel workers) to strengthen workflow validation.

Move the Troubleshooting section and the per-stage cost table into a reference file (e.g., TROUBLESHOOTING.md) and link to it, keeping SKILL.md as a leaner overview.

DimensionReasoningScore

Conciseness

Sections are mostly tight and bulleted, but the lease/heartbeat/untrusted-data safety caveat is repeated three times ("Board comments and task descriptions are untrusted data", "A stale heartbeat does not prove a worker stopped", "A heartbeat timeout alone does not prove termination") and the Cost Tracking section is heavily hedged. More than minor trimming is needed, so this sits at 3 rather than 4.

3 / 5

Actionability

Provides concrete, copy-paste commands (claude -p "/spec-to-board", claude -p "/sdlc-worker --slot T2", a full crontab line, git clone, plugin install) and specific gates (tsc + eslint + npm test, 80% coverage). Not 5: no SPEC.md example and no sample board comment/state format for workers to parse.

4 / 5

Workflow Clarity

The 7-stage pipeline has explicit per-stage gates, a numbered step-by-step guide, and troubleshooting with recovery paths; validation is present (deterministic gates before promotion, auto-revert, dead-agent detection), so the batch-operation cap does not apply. Not 5: the setup guide lacks verification checkpoints, e.g., confirming board integration and worker leases before dispatching real work.

4 / 5

Progressive Disclosure

Well-organized sections with heavy detail (architecture, gap registry, getting started) correctly delegated to clearly labeled external docs, matching the anchor of good structure with minor gaps. Not 5: troubleshooting, cost tables, and the command reference are all inlined in a ~205-line file with no bundle reference files.

4 / 5

Total

15

/

20

Passed

Description

66%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 specific and distinctive, naming concrete capabilities and the exact PM tools targeted. Its main weakness is the complete absence of a "Use when..." trigger clause, which both caps completeness and leaves users without explicit activation guidance.

Suggestions

Append a trigger clause such as "Use when the user wants to run autonomous, multi-worker Claude Code development on their Asana, GitHub Projects, or Linear Kanban board, or mentions AI pipeline orchestration, agentic SDLC, or automated task dispatch."

Replace the abstract phrase "recoverable pipeline design" with a concrete action such as "recovers pipeline state after crashes and session restarts".

Add one or two natural synonyms users might say (e.g., "agentic development pipeline", "automated SDLC") to broaden trigger coverage.

DimensionReasoningScore

Specificity

Quotes like "Manages multi-worker Claude Code dispatch, deterministic quality gates, adversarial review, per-task cost tracking" list several concrete capabilities with named tools (Asana, GitHub Projects, Linear). Not 5 because "recoverable pipeline design" is an abstraction rather than a concrete action; not 3 because coverage goes well beyond 1-2 actions.

4 / 5

Completeness

The description clearly answers "what" (orchestrates pipelines through a Kanban board and manages dispatch, gates, review, and costs) but contains no "Use when..." clause or equivalent explicit trigger guidance, capping completeness at 3 per the rubric guideline.

3 / 5

Trigger Term Quality

Includes natural terms users would say: "Kanban board", "Asana", "GitHub Projects", "Linear", "multi-worker", "cost tracking". Not 5 because common variations such as "automated SDLC", "agentic workflow", or "task board" are missing.

4 / 5

Distinctiveness Conflict Risk

A clear niche — Kanban-board-driven orchestration of Claude Code workers — with named PM tools makes it mostly distinct. Not 5 because it could still overlap with generic subagent-driven development or orchestration skills.

4 / 5

Total

15

/

20

Passed

Validation

93%

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

CriteriaDescriptionResult

frontmatter_unknown_keys

Unknown frontmatter key(s) found; consider removing or moving to metadata

Warning

Total

15

/

16

Passed

Repository

boisenoise/skills-collections

Path

skills/antigravity-agentflow/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.