Tessl • The management layer for AI agent skills

SkillPublic

data-loss-gate

Confirmation gate before any bulk delete, cleanup, or destructive operation that could result in data loss — shell-level (rm -rf, git rm, bulk sed) or brain-level (bulk forget, delete sweeps, purge-deleted, source removal, raw-SQL truncation). Presents a recoverability card and requires an explicit "yes" from the user before proceeding. Routing convention, not an operation-boundary enforcement.

Quality

89%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

Securityby

Passed

No findings from the security scan

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.

An excellent instruction-only body: the workflow is unambiguous with validation checkpoints and error-recovery loops exactly where a destructive-operation skill needs them, and the guidance is copy-paste concrete. The weaknesses are redundancy (the Contract section deliberately duplicates the body, and key purge rules are stated twice) and a monolithic single-file layout that inlines a long command catalog.

Suggestions

Trim the duplication: the Contract section restates the card, logging, and routing guarantees already given in the body, and the purge/`--confirm-destructive` rules appear in both "When This Fires" and "When it fails" — state each once and cross-reference.

Shorten or cut the "Why This Exists" narrative to its one-line rule ("if it's data and it's bulk, ASK FIRST") since the anti-patterns section already carries the motivation.

Consider moving the gbrain-specific command catalog (purge-deleted, source/mount removal, raw-SQL cases) into a reference file so the routing-critical gate procedure stays lean in SKILL.md.

Dimension	Reasoning	Score
Conciseness	The body is dense with non-obvious operational rules and teaches no concepts Claude already knows, but it carries trimmable duplication: the Contract section restates body guarantees ("this section exists for the conformance test"), "Why This Exists" is narrative motivation, and the purge protocol and --confirm-destructive point each appear in two sections.	4 / 5
Actionability	Fully executable guidance throughout: exact commands ("gbrain pages purge-deleted --dry-run --json", "gbrain get <slug> --include-deleted", "gbrain write-request <request_id>"), a copy-paste confirmation card template, a templated deletion-log entry, and explicit thresholds (10 files, >1GB or >1000 files/pages).	5 / 5
Workflow Clarity	Four clearly sequenced steps with explicit validation and feedback loops for a destructive-operation skill: per-page purge verification ("confirm each one is gone… before the next"), response handling that re-presents the card on questions or "wait", write_pending receipt polling before logging, stop-at-first-refusal, and an interrupted-batch protocol, plus the recoverability checklist.	5 / 5
Progressive Disclosure	Clear section structure with well-signaled, one-level-deep references to shared conventions (ask-user "supplies the stop-and-wait discipline"; test-before-bulk "gates bulk WRITE quality") instead of inlining them. However, the ~35-line gbrain command catalog and purge protocol are inlined in a 230-line single-file body with no bundle files to offload detail into.	4 / 5
	Total	18 / 20 Passed

Description

88%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.

A strong description: it states a clear what and an explicit, enumerated when in third person, with concrete trigger phrases and honest scope framing. Its main gaps are a few missing natural synonyms (which only appear in the triggers list) and minor overlap risk with routine-cleanup skills that the description alone doesn't disambiguate.

Suggestions

Fold the natural trigger phrases from the triggers list ("wipe the", "free up space", "drop the table", "truncate") into the description text so users' most common phrasings are matched where routing actually reads.

Add a brief disambiguation clause to the description (e.g. "routine non-destructive cleanup is out of scope") to reduce overlap with maintenance/cleanup skills without relying on the body's dedup section.

Dimension	Reasoning	Score
Specificity	The description lists multiple concrete capabilities — "Presents a recoverability card and requires an explicit 'yes' from the user" — and comprehensively enumerates the gated operations at both levels ("rm -rf, git rm, bulk sed", "bulk forget, delete sweeps, purge-deleted, source removal, raw-SQL truncation"), fully covering the gate's scope.	5 / 5
Completeness	Both what ("Confirmation gate… Presents a recoverability card and requires an explicit 'yes'") and when ("before any bulk delete, cleanup, or destructive operation that could result in data loss", with the shell-level and brain-level operation lists) are explicitly and concretely answered with trigger phrases.	5 / 5
Trigger Term Quality	Good natural-term coverage ("bulk delete", "cleanup", "rm -rf"), but terms like "purge-deleted" and "raw-SQL truncation" are technical jargon, and common natural phrasings users would say ("wipe the", "free up space", "drop the table", "truncate") appear only in the triggers list, not in the description itself.	4 / 5
Distinctiveness Conflict Risk	The pre-destruction confirmation niche is distinct with specific triggers, but "cleanup" and generic destructive-operation phrasing create minor overlap risk with closely related maintenance and decision-gate skills; the dedup boundaries against those siblings live in the body, not the description.	4 / 5
	Total	18 / 20 Passed

Validation

87%

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 — 14 / 16 Passed

Validation for skill structure

Criteria	Description	Result
frontmatter_unknown_keys	Unknown frontmatter key(s) found; consider removing or moving to metadata	Warning
relative_links	Relative link issues: 7 suspicious	Warning

	Total	14 / 16 Passed

Repository: garrytan/gbrain
Path: plugin-variants/gbrain-coding/skills/data-loss-gate/SKILL.md
Revision: a89a5ad

Reviewed: 13 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.

data-loss-gate

Table of Contents

Is this your skill?