CtrlK
BlogDocsLog inGet started
Tessl Logo

continue-claude-code-work

Continues interrupted Claude Code work after local history is read and reconciled — reconstructs the original outcome, unfulfilled requests, and the next action. Use when given a Claude Code Session ID, asked to pick up prior Claude work, told a run was interrupted or compacted, or asked to take over without `claude --resume`. Not for Codex sessions (use continue-codex-work).

74

Quality

93%

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

SKILL.md
Quality
Evals
Security

Quality

Content

85%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 tightly written instruction-only skill: it prescribes a verification-first workflow with concrete tool invocations, an evidence contract, and anti-drift checkpoints. Minor room to improve by adding a filled example contract and trimming a few justificatory sentences.

Suggestions

Include one brief worked example of a completed continuation contract (field -> evidence) so the Step 2 table is unambiguous in practice.

Trim justificatory sentences (e.g. 'continuing from a guessed Session is worse than asking for the missing evidence') to pure directives to gain token efficiency.

Specify the fallback if the history reader Skill has no 'full mode' for clipped sections (e.g. re-run with expanded coverage or stop and ask).

DimensionReasoningScore

Conciseness

The body is dense, directive, and table-driven with essentially no concept re-teaching, but a few rationale sentences ('continuing from a guessed Session is worse than asking for the missing evidence', 'A transcript records attempts as well as successes') are justification rather than instruction and could be trimmed — matching anchor 4 rather than the fully lean anchor 5.

4 / 5

Actionability

Guidance is mostly executable: a named Skill invocation ('Invoke daymade-claude-code:read-claude-code-history'), an evidence-required contract table, numbered reconciliation steps, and explicit prohibitions (Do not run claude --resume). It falls short of anchor 5 only because there is no worked example of a filled-in continuation contract and the fallback when the reader lacks a 'full mode' is left implicit.

4 / 5

Workflow Clarity

Five clearly sequenced steps with explicit validation checkpoints: a read-receipt checklist before acting, stop conditions on unresolved identity, independent verification of prior claims in Step 3, verification of each action's result in Step 4, and a mechanical decision rule ('If this succeeds, which original unfulfilled result becomes smaller?') that functions as a feedback loop against drift.

5 / 5

Progressive Disclosure

The skill is a single well-organized SKILL.md with no bundle files and no need for external references; sections are clearly labeled and the one cross-reference is a one-level-deep Skill invocation, satisfying the no-external-references exception for a well-organized single-file skill.

5 / 5

Total

18

/

20

Passed

Description

100%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.

An exemplary description: concrete capabilities, comprehensive natural trigger phrases, explicit use-when guidance, and a clear disambiguation boundary against a sibling skill. Third-person voice throughout with no padding.

DimensionReasoningScore

Specificity

The description lists multiple concrete actions — 'Continues interrupted Claude Code work after local history is read and reconciled', 'reconstructs the original outcome, unfulfilled requests, and the next action' — giving comprehensive coverage within its domain, matching the anchor-5 pattern rather than the minor-gaps anchor 4.

5 / 5

Completeness

It explicitly answers both 'what' (continues/reconstructs interrupted work) and 'when' via a full 'Use when...' clause with concrete trigger phrases, mirroring the anchor-5 good example.

5 / 5

Trigger Term Quality

Trigger phrases are comprehensive and natural: 'given a Claude Code Session ID, asked to pick up prior Claude work, told a run was interrupted or compacted, or asked to take over without claude --resume' covers the ways a user would actually phrase this need.

5 / 5

Distinctiveness Conflict Risk

It carves a clear niche (interrupted Claude Code sessions) and adds an explicit boundary — 'Not for Codex sessions (use continue-codex-work)' — minimizing conflict risk with sibling skills.

5 / 5

Total

20

/

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
daymade/claude-code-skills
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.