Tessl • The management layer for AI agent skills

The Membrane CLI mechanics — install, headless auth, connection lifecycle with polling, action discovery/run, and proxy fallback — are genuinely actionable and well sequenced. But the file is marred by leftover generation artifacts (a placeholder sentence, an Apple Bluetooth docs link that has nothing to do with the Blue app), an inconsistent entity model (cases/contacts vs. todos/projects/companies) that contradicts the action table, broken step references, and no validation guidance for destructive actions.

Suggestions

Delete the placeholder line "I don't have enough information to describe this app..." and the irrelevant Apple Bluetooth documentation link — both are generation artifacts that waste tokens and mislead.

Reconcile the entity model: the bullet dump lists Case/Contact/Case Tag entities while the Popular actions table exposes todos/projects/companies — align them, and move the full entity/action catalog to a references/ file with a concise inline summary.

Fix workflow navigation (label the steps so 'skip to Step 2' and '1b' resolve) and add validation guidance for destructive operations, e.g. 'Before delete-todo or set-todo-assignees, show the current record via get-todo and confirm with the user.'

Dimension	Reasoning	Score
Conciseness	Noticeably verbose with padded sections that earn no tokens: the placeholder line "I don't have enough information to describe this app. Please provide more details...", the irrelevant "Official docs: https://developer.apple.com/documentation/bluetooth" link, and a 37-line entity bullet dump capped only by "Use action names and parameters as needed." Not a 1 because the CLI command sections themselves are lean and free of basic-concept explanations.	2 / 5
Actionability	Mostly executable guidance with copy-paste commands throughout: "npm install -g @membranehq/cli@latest", "membrane connection ensure \"https://www.blue.cc/\" --json", "membrane action run <actionId> --connectionId=CONNECTION_ID --input '{\"key\": \"value\"}' --json", plus a flags table for proxy requests. Not a 5 because no worked example uses real parameter values (e.g. an actual create-todo input), and 'Use action names and parameters as needed' is a vague instruction rather than an executable step.	4 / 5
Workflow Clarity	A real sequence exists (install → login → connection ensure → poll until READY → search actions → run), with state-by-state handling (READY / CLIENT_ACTION_REQUIRED / CONFIGURATION_ERROR), but validation is absent around destructive operations: the action table includes Delete Todo and replace-style updates (Set Todo Assignees 'replaces existing assignees') with no verify-before/after guidance, which caps workflow clarity at 3. Navigation is also broken — "skip to Step 2" references a step that is never labeled, and "#### 1b" appears with no 1a.	3 / 5
Progressive Disclosure	No bundle files exist (no references/, scripts/, or assets/), so this scores on the body's own structure: headers, tables, and code blocks are present, but organization is mediocre — heading hierarchy is inconsistent (h3 'Running actions' and 'Proxy requests' fall under h2 'Popular actions'), orphan placeholder text sits at the top, and the entity dump and 19-row action table are inlined reference material that could live in a separate file. Fits 'Some structure but could be better organized' rather than the well-placed content of a 4.	3 / 5
	Total	12 / 20 Passed

Description

48%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 correctly names the target app and includes a 'Use when' clause, but the capability statement is generic filler that never says what Blue integration actually does (manage projects, todos, records, comments). Trigger coverage is thin — none of the natural user vocabulary for a project-management tool appears.

Suggestions

Replace the generic tail with concrete capabilities, e.g. 'Create, update, and delete projects, todos (records), todo lists, tags, and comments in Blue; manage assignees and companies; run bulk operations via the Membrane CLI.'

Expand the 'Use when' clause with natural trigger phrases users would actually say, e.g. 'Use when the user mentions Blue, blue.cc, or asks to manage Blue projects, todos, records, tags, or comments.'

Mention key synonyms (todos/tasks/records; companies/workspaces) so the description fires on the varied vocabulary Blue users use.

Dimension	Reasoning	Score
Specificity	"Manage data, records, and automate workflows" names the Blue domain but the actions are entirely generic — no concrete capabilities like projects, todos, comments, or record CRUD are mentioned. This matches the anchor 'Names the domain but actions are minimal or generic'; a 3 would require at least 1-2 concrete actions.	2 / 5
Completeness	A "when" is explicitly present ("Use when the user wants to interact with Blue data") but the "what" ("Manage data, records, and automate workflows") is vague, and the when-clause merely restates the domain rather than giving concrete triggers. It falls between anchor 3 (clear what, weak when) and 4 (both present, when could be more specific) — the vagueness of the what keeps it at 3.	3 / 5
Trigger Term Quality	"Blue", "records", and "Blue data" are relevant keywords, but common natural variations a Blue user would say (todos, tasks, projects, comments, companies) are missing. Fits 'Some relevant keywords but missing common variations or synonyms' rather than the comprehensive coverage of a 4.	3 / 5
Distinctiveness Conflict Risk	"Blue integration" names a specific app, making it mostly distinct with minor overlap risk from the generic "data, records, workflows" terms. Not a 5 because "Blue" is also a common English word and the generic tail could still collide with other data-management skills.	4 / 5
	Total	12 / 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
frontmatter_unknown_keys	Unknown frontmatter key(s) found; consider removing or moving to metadata	Warning

	Total	15 / 16 Passed

Repository: membranedev/application-skills
Path: skills/blue/SKILL.md
Revision: f484c82

Reviewed: 16 days ago

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.

blue

Table of Contents

Is this your skill?