CtrlK
BlogDocsLog inGet started
Tessl Logo

ingest

Route content to specialized ingestion skills. Detects input type and delegates.

48

Quality

53%

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 ./skills/ingest/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.

A substantive, largely actionable instruction body with sequenced phases, a real batch-validation loop, and concrete gbrain commands. Its weaknesses are repetition of the same mandates across five sections, and a monolithic single-file layout whose key referenced files are missing from the bundle.

Suggestions

State the Iron Law / back-link rule once (e.g., in the Contract) and reference it from Phases, Quality Rules, and Anti-Patterns instead of restating it verbatim five times; likewise deduplicate the citation and raw-source-preservation mandates.

Move the per-format media workflows (articles, videos/podcasts, PDFs, screenshots, meetings, social) into a bundled reference file (e.g., `references/media-workflows.md`) and keep one-line summaries with links in SKILL.md, so the overview stays lean.

Bundle `skills/_brain-filing-rules.md` (or inline its notability gate and filing directory map), since multiple mandated steps currently depend on a file the skill does not ship.

DimensionReasoningScore

Conciseness

The body is dense with domain-specific rules rather than explanations of known concepts, but the same rules are repeated heavily: "An unlinked mention is a broken brain" appears verbatim in both the Contract and Anti-Patterns sections, the back-link/Iron Law rule is restated roughly five times, citation requirements are stated twice, and "Test Before Bulk" ends with rhetorical padding ("The marginal cost of testing 3 items first is near zero..."). This matches the anchor for mostly efficient content that could be tightened; it is above anchor 2 because the bulk is substantive, not filler.

3 / 5

Actionability

Concrete commands and formats are given throughout: `gbrain search "name"`, `gbrain put <type/slug>`, `gbrain sync --no-pull --no-embed`, `gbrain files upload-raw <file> --page <page-slug> --type <type>`, explicit `[Source: ...]` citation templates, a `.redirect.yaml` example, and a copy-paste output format block. Minor gaps keep it below 5: "Fetch content (`web_fetch` or equivalent)" is ambiguous, and the notability gate defers to a filing-rules file not present in the bundle.

4 / 5

Workflow Clarity

Phases 1-6 are explicitly numbered, the entity-detection protocol is stepwise, and the "Test Before Bulk" section supplies a genuine validate -> read output -> fix -> retry loop for batch operations, so the batch-validation cap does not apply. It sits at anchor 4 rather than 5 because individual write operations (back-link creation, entity page updates) have no verification checkpoint.

4 / 5

Progressive Disclosure

Sections are well organized and references are clearly signaled (the top filing-rule blockquote and inline mentions), but the bundle contains no reference files — the repeatedly cited `skills/_brain-filing-rules.md` and `skills/conventions/quality.md` are not present — and roughly 100 lines of per-format media workflows plus the citation format spec are inlined in a single ~290-line file. This matches the anchor for some structure with content that should be separate left inline; not 4 because the load-bearing referenced material is unavailable and split-out candidates remain inline.

3 / 5

Total

14

/

20

Passed

Description

42%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 communicates the skill's router role but is thin: two actions, almost no natural trigger language, and no "when to use" guidance. Most of the useful trigger phrasing lives in the frontmatter triggers rather than the description field where it would aid discovery.

Suggestions

Add an explicit "Use when..." clause, e.g., "Use when the user says 'ingest this', 'save this to brain', or shares an article, meeting transcript, video, or document to capture."

Include natural trigger vocabulary in the description itself — article, meeting, podcast, transcript, screenshot, "save this" — rather than only the technical term "ingestion".

Name the delegate targets (idea-ingest, media-ingest, meeting-ingestion) and one or two concrete ingestion actions (create entity pages, back-link mentions, append timeline entries) so the "what" is comprehensive rather than minimal.

DimensionReasoningScore

Specificity

"Route content to specialized ingestion skills. Detects input type and delegates." names the domain and exactly two concrete actions but does not enumerate content types or what the specialized ingestion skills do. It matches the anchor for 1-2 concrete actions without comprehensive coverage; it is above anchor 2 because the actions are more specific than a generic one-liner.

3 / 5

Completeness

The "what" is reasonably clear (a router that detects input type and delegates to specialized ingestion skills), but there is no "Use when..." clause or equivalent trigger guidance, which caps completeness at 3 per the judging guidelines. It is not 4 because the "when" is entirely absent rather than merely implicit or under-specified.

3 / 5

Trigger Term Quality

The description's only user-sayable keyword is "ingestion"; natural phrases a user would actually say (e.g., "save this", "process this meeting", "article", "meeting notes") appear only in the frontmatter triggers, not the description. This matches the anchor for one or two generic keywords missing the natural phrases; it is below anchor 3 because there are not multiple relevant keywords present.

2 / 5

Distinctiveness Conflict Risk

Signaling that it routes to sibling ingestion skills helps distinguish it, but "content" is broad and it would overlap with the very specialized skills it names (idea-ingest, media-ingest, meeting-ingestion). This matches the anchor for somewhat specific with residual overlap risk; not 4 because the overlap with its own delegate targets is real.

3 / 5

Total

11

/

20

Passed

Validation

87%

Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.

Validation — 14 / 16 Passed

Validation for skill structure

CriteriaDescriptionResult

frontmatter_unknown_keys

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

Warning

relative_links

Relative link issues: 1 missing

Warning

Total

14

/

16

Passed

Repository
garrytan/gbrain
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.