CtrlK
BlogDocsLog inGet started
Tessl Logo
SkillPublic

dsh-config-sync

把 DSH(DeepSeek Harness)的用户配置(~/.dsh 下的 AGENTS.md、settings.yaml 及可选 profiles/预设)打包成脱敏骨架并在设备间安全同步恢复,提供只读审计、显式应用与 SHA-256 校验,绝不携带凭据和运行态。用于复现 DSH 环境、跨设备迁移偏好与模型路由配置。

Quality

79%

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 ./skills/dsh-config-sync/SKILL.md

The canonical home for this skill is dsh-config-sync in ooooooooooooooooooop/agent-tools

SKILL.md
Quality
Evals
Security

Quality

Content

66%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 is well-structured and secure by design, with an exemplary validated export/restore workflow (scan → FAIL/stop, check → apply → re-verify). Its main weaknesses are that it never surfaces the bundle — no pointer to references/manifest.md or the sync_dsh_config.py script — leaving the guidance not directly executable.

Suggestions

Add the concrete invocation for the bundled tool, e.g. 'python scripts/sync_dsh_config.py --check / --apply / export [--template] [--with-optional]', in the 打包 and 恢复 sections so the named modes and flags are executable.

Replace or trim the inline 纳入/排除 table and link to references/manifest.md (e.g. '完整清单与敏感扫描规则见 [manifest.md](references/manifest.md)') to remove the duplication and surface the bundle file.

Consolidate the forced-exclusion rules (.credentials.yaml / sessions / storages) into one authoritative place (the table or 安全边界) to cut the repetition across three sections.

DimensionReasoningScore

Conciseness

The body is dense and assumes competence — tables for include/exclude and template placeholders, no explanation of what SHA-256 or sanitization is. It is not a 5 because the exclusion rules are stated three times (table '强制排除', the 安全边界 section, and export step 2) and the inline 纳入/排除 table duplicates references/manifest.md, so minor trimming is possible.

4 / 5

Actionability

Modes and flags are named in prose ('--template', '--with-optional', check/apply/export, SHA-256, manifest.json), but the body never shows how to invoke the tool — scripts/sync_dsh_config.py is never mentioned and there is no command line anywhere. This is concrete-but-incomplete guidance with missing key execution details, matching anchor 3 rather than anchor 4's 'concrete code or commands'.

3 / 5

Workflow Clarity

Export (steps 1–7) and restore (steps 1–4) are clearly sequenced with explicit validation checkpoints and feedback loops: sensitive scan after copy marks the package FAIL and stops; restore is read-only check first, then explicit apply, then a re-check with SHA-256 verification; PASS requires clean post-apply verification, with PARTIAL for non-blocking diffs. This matches anchor 5 exactly, and the destructive/batch cap does not apply since validation is present.

5 / 5

Progressive Disclosure

Two bundle files exist (references/manifest.md, scripts/sync_dsh_config.py) but the body references neither: the 纳入/排除 table inlines what references/manifest.md already covers instead of pointing to it, and the flags imply a script that is never linked. Section structure is good, so it is above anchor 2's 'minimal structure', but the references are effectively buried/absent, keeping it below anchor 3's 'references present but not clearly signaled'.

2 / 5

Total

14

/

20

Passed

Description

83%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: concrete, niche-scoped, third-person, and explicit about both capabilities and use situations, with clear safety boundaries (no credentials, no runtime state). The only weakness is that the trigger phrasing leans on purpose statements rather than a fuller set of natural trigger variations.

DimensionReasoningScore

Specificity

The description lists multiple concrete actions with named artifacts: '打包成脱敏骨架', '安全同步恢复', '只读审计', '显式应用', 'SHA-256 校验', and names the exact files (~/.dsh 下的 AGENTS.md、settings.yaml 及可选 profiles/预设). Coverage is comprehensive for the skill's scope.

5 / 5

Completeness

Both 'what' (packaging sanitized skeletons, sync/restore, SHA-256 verification, no credentials) and 'when' ('用于复现 DSH 环境、跨设备迁移偏好与模型路由配置') are explicitly answered, so it clears the cap-3 rule. The 'when' is phrased as a purpose clause rather than an explicit 'Use when...' trigger statement, matching anchor 4 ('when' could be more explicit) rather than anchor 5.

4 / 5

Trigger Term Quality

Natural phrases users would say are present: '同步恢复', '跨设备迁移偏好与模型路由配置', '复现 DSH 环境'. A few natural variations are missing (e.g., '备份 DSH 配置', English terms, or the literal path '.dsh'), so it sits between the good-coverage (4) and comprehensive-synonym (5) anchors.

4 / 5

Distinctiveness Conflict Risk

It is tightly scoped to DSH (~/.dsh) user configuration only, explicitly excludes credentials and runtime state, and the domain ('DSH(DeepSeek Harness)的用户配置') is a clear niche with distinct triggers, giving minimal conflict risk with other sync/config skills.

5 / 5

Total

18

/

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

ooooooooooooooooooop/personal-ai

Path

skills/dsh-config-sync/SKILL.md

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.