transcript-saver
Save and export the current Claude Code session as a shareable HTML transcript. Use this skill when the user asks to save, export, archive, or publish their current Claude Code conversation or session. Triggers on phrases like "save this transcript", "export this session", "create a transcript", "archive this conversation", "publish to gist", or "share this session". Wraps Simon Willison's claude-code-transcripts tool for in-session use.
85
82%
Does it follow best practices?
Impact
85%
2.83xAverage score across 3 eval scenarios
Critical
Do not install without reviewing
Quality
Discovery
100%Based on the skill's description, can an agent find and select it at the right time? Clear, specific descriptions lead to better discovery.
This is an excellent skill description that clearly communicates what the skill does (exports Claude Code sessions as HTML transcripts), when to use it (with explicit trigger phrases), and how it works (wraps Simon Willison's tool). It follows all best practices: third person voice, explicit 'Use when' clause, natural trigger terms, and a distinct niche that won't conflict with other skills.
| Dimension | Reasoning | Score |
|---|---|---|
Specificity | Lists specific concrete actions: 'save and export the current Claude Code session as a shareable HTML transcript'. Also mentions wrapping Simon Willison's claude-code-transcripts tool, which is a concrete implementation detail. | 3 / 3 |
Completeness | Clearly answers both 'what' (save/export Claude Code session as shareable HTML transcript) and 'when' (explicit 'Use this skill when...' clause with detailed trigger phrases). The 'Triggers on phrases like...' section provides additional explicit guidance. | 3 / 3 |
Trigger Term Quality | Excellent coverage of natural trigger terms: 'save', 'export', 'archive', 'publish', 'transcript', 'session', 'conversation', 'share', 'publish to gist'. These are terms users would naturally use when wanting this functionality. | 3 / 3 |
Distinctiveness Conflict Risk | Highly distinctive - specifically targets Claude Code session transcripts, mentions HTML format, references a specific tool (claude-code-transcripts), and the trigger terms are narrowly scoped to session saving/exporting rather than general file operations. | 3 / 3 |
Total | 12 / 12 Passed |
Implementation
64%Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.
This skill provides clear, actionable commands for saving Claude Code transcripts with good coverage of options and troubleshooting. Its main weaknesses are moderate verbosity (explaining internals and concepts Claude doesn't need) and lack of output verification steps. The content would benefit from trimming explanatory sections and adding a brief validation checkpoint after export.
Suggestions
Remove or significantly trim the 'How It Works' section — Claude doesn't need to know the internal pipeline; just the commands and their effects.
Add a brief verification step after running the script, e.g., 'Verify output: check that index.html and at least one page-NNN.html exist in the output directory.'
Move Installation Requirements and Troubleshooting to a separate REFERENCE.md or TROUBLESHOOTING.md file, keeping SKILL.md focused on the quick start and command reference.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The content includes some unnecessary explanation (e.g., 'How It Works' section explaining what the tool does internally, 'Important Notes' about session timing that Claude can infer). The installation requirements section explains what uv and gh are, which is padding. However, the command reference and examples are reasonably tight. | 2 / 3 |
Actionability | Provides fully executable bash commands with clear flags and options. The command reference is complete, examples are copy-paste ready with expected output shown (e.g., the gist preview URL), and troubleshooting includes specific fix commands. | 3 / 3 |
Workflow Clarity | This is essentially a single-command tool, so multi-step workflows aren't deeply needed. However, there's no validation step mentioned — no guidance on verifying the output HTML was generated correctly, checking file counts, or confirming gist upload succeeded. For a tool that publishes to external services, some verification guidance would be appropriate. | 2 / 3 |
Progressive Disclosure | Content is reasonably structured with clear sections, but it's somewhat monolithic — the troubleshooting, installation requirements, and detailed 'How It Works' sections could be in a separate reference file. For a skill that's essentially 'run this one script with flags,' there's a lot of inline content that could be offloaded. | 2 / 3 |
Total | 9 / 12 Passed |
Validation
100%Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.
Validation — 11 / 11 Passed
Validation for skill structure
No warnings or errors.
- Repository
- fpl9000/ai-skills
- Commit
3f74412
- 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.