session-historian

Look up prior Claude Code sessions when context is lost or forgotten. Use when asked what was done before, what happened in the last session, or any request to recall past conversation history, prior decisions, experiments, or outcomes. Indexes and searches raw JSONL transcripts from ~/.claude/projects/ via DuckDB. Returns verbatim user messages, summarizes AI actions and sub-agent outcomes, detects tool errors and user frustration signals, and reports tool usage statistics. Summaries cached at ~/.claude/kaizen/session-summaries/.

Quality

91%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

Securityby

High

Do not use without reviewing

Quality

Content

82%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 a well-structured, highly actionable reference with concrete commands and a clear recall workflow; its main weakness is mild verbosity from implementation-detail paragraphs and the inlined command/schema reference that could live in separate files. No destructive/batch validation cap applies.

Suggestions

Trim implementation-detail prose like the "two-pass scan" explanations for errors/tools — Claude only needs the command, flags, and output format, not the internal algorithm.

Move the full Command Reference and JSONL Schema Reference into a separate reference file (e.g. references/commands.md) and link to it from SKILL.md to improve progressive disclosure and reduce token load.

Replace the example-filled entries inside the Summary Template with bare placeholders to cut tokens while keeping the structure clear.

Dimension	Reasoning	Score
Conciseness	Mostly efficient with concrete commands and examples, but includes some implementation-detail prose (e.g. "Performs a two-pass scan: first builds a map of tool-use IDs...") and an example-filled summary template that could be trimmed; not score 5 because those passages explain mechanics Claude does not strictly need, and not score 3 because the bulk is lean and command-oriented.	4 / 5
Actionability	Fully executable, copy-paste-ready commands throughout (list/messages/search/show/index/errors/tools/irritation/current-path) with exact flags, raw output formats, and exit codes, plus a numbered workflow with concrete invocations; not score 4 because the guidance covers the common cases completely rather than having minor gaps.	5 / 5
Workflow Clarity	The "I forgot what happened" workflow is a clear 4-step sequence with concrete commands, and Fidelity Rules act as a quality checklist; not score 5 because there is no explicit validate-then-retry feedback loop, and not score 3 because steps are well defined and the destructive/batch cap does not apply to these read-mostly operations.	4 / 5
Progressive Disclosure	Good structure with clear section headers and a single one-level-deep script reference (the bundle contains only scripts/session_query.py, no nested references); not score 5 because the sizable Command Reference / JSONL Schema Reference are inlined rather than split into reference files, and not score 3 because navigation is clear and references are well signaled rather than buried.	4 / 5
	Total	17 / 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.

The description is strong across all dimensions: it states concrete capabilities, gives explicit "Use when..." trigger guidance with natural phrasing, and carves out a distinct niche. Third-person voice is maintained throughout, consistent with the rubric's voice guidance.

Dimension	Reasoning	Score
Specificity	Lists multiple concrete actions — "Indexes and searches raw JSONL transcripts", "Returns verbatim user messages", "summarizes AI actions and sub-agent outcomes", "detects tool errors and user frustration signals", "reports tool usage statistics" — giving comprehensive coverage; not score 4 because the action list is broad rather than having only minor gaps.	5 / 5
Completeness	Explicitly answers both: "what" (indexes/searches, returns verbatim messages, summarizes, detects errors, reports stats) and "when" ("Use when asked what was done before, what happened in the last session, or any request to recall past conversation history..."); not score 4 because the trigger guidance is concrete and explicit, not merely present.	5 / 5
Trigger Term Quality	Natural user phrases are well covered — "what was done before", "what happened in the last session", "recall past conversation history, prior decisions, experiments, or outcomes" — matching how a user would actually ask; not score 4 because these cover the common variations users say.	5 / 5
Distinctiveness Conflict Risk	Clear niche — recalling prior Claude Code sessions via JSONL transcripts/DuckDB — with distinct triggers unlikely to fire for unrelated skills; not score 4 because the domain and triggers are tightly scoped with minimal overlap risk.	5 / 5
	Total	20 / 20 Passed

Validation

100%

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 — 16 / 16 Passed

Validation for skill structure

No warnings or errors.

Repository: Jamie-BitFlight/claude_skills
Path: .claude/skills/session-historian/SKILL.md
Commit: c337daf

Reviewed: 3 days ago

Table of Contents

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.