CtrlK
BlogDocsLog inGet started
Tessl Logo

archive-protocol

全局归档协议。只要任务需要写入任何文件(含 PLAN.md、报告、JSON 等归档物),必须按本技能执行 Session→ARCHIVE_ID、TIMESTAMP、双轨路径(根段须为 archives/)、回读校验与状态回执;WebUI 的 archive_grid 必须用 Markdown 中语言标识为 json 的围栏代码块输出。

58

Quality

67%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

SecuritybySnyk

Passed

No findings from the security scan

Fix and improve this skill with Tessl

tessl review fix ./configs/microservice/bff-service/configs/agent-skills/ontology/archive-protocol/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 delivers a concrete, actionable archiving protocol with real tool-call examples, explicit post-write verification, and clear error-handling guidance. Its weaknesses are redundancy — repeated path templates, duplicated dual-track rules, identical JSON examples, and a tangential identity-file section — and a workflow spread across overlapping sections rather than one consolidated validated sequence.

Suggestions

Consolidate the repeated absolute-path template and the dual-track rules (stated in both 【归档路径规则】 and 【工具执行指引】) into a single section, and merge the two identical JSON examples into one.

Remove or relocate the tangential 【身份文件】 (SOUL.md/IDENTITY.md) section, which is unrelated to the archive workflow.

Present one consolidated end-to-end workflow (prepare → call archive → verify read-back → output card) with validation as an explicit checkpoint step, and include an example of the required markdown json-code-block output card.

DimensionReasoningScore

Conciseness

The body is rule-dense with no explanations of known concepts, but includes several padded/redundant sections: the absolute path template is repeated about five times, dual-track rules appear in both 【归档路径规则】 and 【工具执行指引】 with two literally identical JSON examples, and the 【身份文件】 section on SOUL.md/IDENTITY.md is tangential to archiving. This sits between 'noticeably verbose' (2) and 'efficient with minor trimming' (4).

3 / 5

Actionability

Concrete, copy-pasteable guidance is provided: exact tool-call JSON ("{\"kind\":\"file\",\"sourcePath\":\"result.json\"}"), the plan-file parameter shape, exact absolute path templates, and explicit 'Archive blocked' error handling. Minor gaps keep it below 5: the two JSON examples (single file vs directory) are identical payloads, and the required final markdown json-code-block output card has no example.

4 / 5

Workflow Clarity

Clear numbered sequences exist (【目录归档模式】's 4-step folder flow and 【工具执行指引】's prepare→call→verify→handle-error flow) with an explicit validation mandate ("任何归档文件写入后,必须立即回读校验。未完成校验前,不得声称「已成功归档」") and an error-recovery loop for 'Archive blocked'. Not 5 because validation is partly delegated to the tool's automatic read-back rather than given as an explicit checkpoint step, and the workflow is scattered across overlapping sections instead of one consolidated sequence.

4 / 5

Progressive Disclosure

No bundle files exist and none are needed; the body is self-contained with clear 【】-labeled sections covering boundary, path rules, directory mode, verification, and tool usage. Good structure with minor organization gaps (overlapping sections could be consolidated); the under-50-line exception to score 5 does not apply at ~95 lines.

4 / 5

Total

15

/

20

Passed

Description

67%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 explicitly answers both what the skill does and when to use it, with recognizable trigger terms. Its main weakness is an extremely broad trigger condition — any file write — that risks conflicts with nearly every file-producing skill, alongside dense internal jargon (ARCHIVE_ID, 双轨路径, archive_grid) that limits natural-term coverage.

Suggestions

Narrow the 'when' clause from 'any task that writes any file' to concrete archival moments (e.g., 'Use when finishing a task that produced deliverables like reports, results, or plans that must be archived').

Add natural-language trigger synonyms users would actually say (e.g., 'save results', 'archive output', 'deliverables', 'write-up') alongside PLAN.md/JSON.

Move internal jargon (Session→ARCHIVE_ID, TIMESTAMP, 双轨路径, archive_grid) out of the description and into the body so the frontmatter stays a plain-language what/when summary.

DimensionReasoningScore

Specificity

The description lists several concrete operations — "Session→ARCHIVE_ID、TIMESTAMP、双轨路径(根段须为 archives/)、回读校验与状态回执" plus the json-fenced-code-block rule for WebUI archive_grid — matching 'several specific actions; minor gaps'. It is not 3 (well beyond 1-2 actions) nor 5 (heavy internal jargon and uneven coverage keeps it from comprehensive, plainly-stated coverage).

4 / 5

Completeness

Both 'what' (global archive protocol executing ARCHIVE_ID/TIMESTAMP/dual-track paths/read-back verification/status receipt) and 'when' ("只要任务需要写入任何文件") are explicitly stated. Not 5 because the 'when' is a single broad condition rather than concrete natural trigger phrases; clearly above 3 since 'when' is explicit, not merely implied.

4 / 5

Trigger Term Quality

Natural trigger terms are present: "写入任何文件", "PLAN.md", "报告", "JSON", "归档" — phrases a user would plausibly say. Missing common synonyms and variations (e.g., 'save', 'output files', 'deliverables'), so it falls just below anchor 5.

4 / 5

Distinctiveness Conflict Risk

The trigger "只要任务需要写入任何文件(含 PLAN.md、报告、JSON 等归档物)" fires on virtually any task that produces a file, creating very high overlap risk with any file-producing skill. It is above 1 because the protocol's purpose (archiving) is itself a distinct niche rather than an entirely generic description.

2 / 5

Total

14

/

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
UnicomAI/wanwu
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.