Tessl • The management layer for AI agent skills

A solid, highly actionable CLI-integration skill: commands are concrete and executable, and connection lifecycle handling is unusually thorough. The main gaps are the missing validation/verification guidance around its destructive batch operations (delete-rows, delete-doc) and a dangling 'Step 2' reference to a heading that doesn't exist.

Suggestions

Add a validation step before destructive/batch actions, e.g. 'Before delete-rows/delete-doc, run get-row/get-doc to confirm the target IDs, and confirm with the user for multi-row deletes', to lift workflow clarity past the destructive-operation cap.

Fix the dangling reference by labeling the 'Searching for actions' section as 'Step 2' (or rewording the READY-state line), and state explicitly that the catalog's 'Key' value is what `<actionId>` expects in `membrane action run`.

Move the full action catalog table into a reference file (e.g. references/actions.md) and keep a handful of representative actions in SKILL.md, since actions are also discoverable via `membrane action list`.

Dimension	Reasoning	Score
Conciseness	The body is dominated by executable commands, flag tables, and terse state-handling lists, with only minor padding such as "It's used by teams to centralize information..." and the filler line "Use action names and parameters as needed." Fits 'Efficient; minor instances of over-explanation that could be trimmed' rather than 5's every-token-earns-its-place.	4 / 5
Actionability	Nearly all guidance is copy-paste-ready (install, login, `membrane connection ensure "https://coda.io" --json`, action run with `--input`, the proxy flag table), which approaches 5, but there are minor gaps: the text says "skip to Step 2" yet no step is labeled 'Step 2', and the action catalog's 'Key' column is never explicitly tied to the `<actionId>` the run command expects. Not 3, since everything shown is executable, not pseudocode.	4 / 5
Workflow Clarity	A clear install → login → connect → search → run sequence exists with state-based branching (READY/BUILDING/CLIENT_ACTION_REQUIRED) and explicit polling, but the skill exposes batch and destructive operations (delete-rows, delete-doc) with no validation or verification guidance such as fetching rows first or confirming the target before deleting. Per the rubric, missing validation for destructive/batch operations caps this at 3, which the sequence would otherwise exceed.	3 / 5
Progressive Disclosure	The single file is well organized with clear section headers and a logical flow, no nested references, and dynamic discovery via `membrane action list` offsets the need for external files. Falls short of 5 because the 20-row action catalog table is reference material inlined in SKILL.md that could live in a separate file; above 3 because structure and navigation are good and the inline content is compact, not a wall of API reference.	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 serviceable: it names the product, gives a what and an explicit when, and is safely distinctive. Its main weakness is thin trigger vocabulary — beyond the word 'Coda' it relies on generic phrasing and omits the object-level terms (docs, tables, rows, pages) users would naturally say.

Suggestions

Add concrete actions and object nouns to the what-clause, e.g. 'Create and update docs, tables, rows, and pages; list and filter records; automate workflows' instead of the generic 'Manage data, records'.

Expand the when-clause with natural trigger phrases such as 'Use when the user mentions Coda, a Coda doc, or wants to read, insert, or update rows in a Coda table.'

Dimension	Reasoning	Score
Specificity	The description names the domain objects ("Manage data, records") and actions ("automate workflows"), but the verbs are generic and there are no concrete operations like 'insert rows' or 'create docs'. This matches the anchor 'Names domain and 1-2 concrete actions, but not comprehensive' rather than 4's 'several specific actions' and better than 2, since actions are listed rather than minimal.	3 / 5
Completeness	An explicit when-clause is present ("Use when the user wants to interact with Coda data") alongside a what ("Manage data, records, and automate workflows"), satisfying both halves. Not 5 because the when is broad ('interact with Coda data') rather than concrete trigger phrases, and the what itself is vague; not 3 because the when is explicit, not weakly implied.	4 / 5
Trigger Term Quality	The effective trigger vocabulary is essentially just 'Coda' plus generic terms ('data', 'workflows'), missing the natural Coda nouns users actually say — 'docs', 'tables', 'rows', 'pages'. Fits 'Some relevant keywords but missing common variations or synonyms', short of 4's good keyword coverage.	3 / 5
Distinctiveness Conflict Risk	"Coda integration" tied to "Use when the user wants to interact with Coda data" defines a clear single-product niche with a distinct product-name trigger, minimizing conflict with other skills. This matches the 5 anchor's named-domain distinct triggers.	5 / 5
	Total	15 / 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/coda/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.

coda

Table of Contents

Is this your skill?