read-claude-code-history

An agent skill by daymade, from daymade/claude-code-skills. Tags: analysis, documentation, history, local-workflow, search.

What it does

Reads, searches, and exports local Claude Code history without resuming work. Covers recent session inventory, exact session timelines, verbatim human input including queued mid-turn prompts, full-event keyword search, hybrid recall when wording changed, end-state triage, and deleted-file recovery across active Claude homes plus registered archives. Use whenever the user asks what they or Claude said, wants a Claude Code session ID or original context, remembers prior work vaguely, needs an old file from a transcript, or must prove what a Claude session contained before continuing it. Also owns the only Kimi CLI surface, via its Kimi inventory and search flags. For Codex history use read-codex-history; when the request names no platform at all or spans providers, start at local-conversation-history.

Install

With the skills CLI, which installs into Claude Code, Codex, Cursor and other agents:

npx skills add daymade/claude-code-skills --skill read-claude-code-history

Or copy the skill folder into Claude Code's skills directory by hand (~/.claude/skills for every project, or .claude/skills inside one):

git clone --depth 1 https://github.com/daymade/claude-code-skills
cp -r claude-code-skills/daymade-claude-code/read-claude-code-history ~/.claude/skills/read-claude-code-history

Safety box score

Grade D, elevated risk: 37 of 100 risk points, rated 2026-10-19. The grade rates what the skill and its scripts can reach on the machine of whoever installs it, across eight categories from shell execution to secrets access. How the score works.

This skill reads, searches, exports, and sometimes recovers local Claude Code/Kimi/Codex history from the user's home directories and related scratchpad/archive paths. The main risks are exposure of private prompts, tool calls, hook runs, and other secrets, plus file writes during recovery to destinations outside the project tree.

Category levels, from 0 (none) to 3 (high)
Category	Level	Why
Code & shell execution	2 (moderate)	It ships and runs local Python tools (for example `python3 scripts/read_claude_session.py` / `history_index.py` / `analyze_sessions.py`), and the inspected tests show subprocess use such as `subprocess.run` and `Popen`-based prefiltering, so it can execute local helper processes.
Network access	0 (none)	No upload/download or remote-host communication is evident in the inspected files; the only apparent import hint (`urllib.parse.quote`) is not network I/O.
Filesystem reach	3 (high)	The skill intentionally traverses and reads user-home stores and scratchpads like `~/.claude/`, `~/.claude/history-sources.json`, `/private/tmp/claude-$(id -u)/**/scratchpad/`, `~/.claude-history-index/skill-usage-v1.db`, `~/.codex`, and `~/.kimi-code`, and recovery mode can write output files.
Secrets & credentials	3 (high)	Its purpose is to expose verbatim history, including "user input", tool calls, hook runs, session IDs, and proof of what a session contained; the docs explicitly warn that raw history can contain credentials and private business context.
Privilege & persistence	0 (none)	There is no sudo, service registration, shell-startup editing, or global-install behavior in the inspected skill files; it stays at normal user privilege.
Instruction hijack surface	0 (none)	I found no hidden text, remote instruction fetching, or "ignore your rules"-style prompt injection; the strong directives are the skill's own workflow rules, not untrusted instructions.
Irreversible actions	0 (none)	The skill is read/inspect/export oriented and does not itself push, post, deploy, pay, book, or delete on the user's behalf.
Transparency	1 (low)	Several shipped implementation/reference files were not readable in this review (`scripts/analyze_sessions.py`, `scripts/read_claude_session.py`, `scripts/recover_content.py`, `scripts/history_index.py`, `scripts/_core/text.py`, and some references), so the exact behavior of the main readers/indexers is only partially visible here.

Before installing

Inspect `scripts/recover_content.py` and confirm you are comfortable with it writing recovered files to a separate destination.
Inspect `scripts/read_claude_session.py` and `scripts/analyze_sessions.py` for any additional local subprocess or filesystem reach before using them on sensitive history.
Check whether `~/.claude/history-sources.json`, `~/.claude-history-index/skill-usage-v1.db`, and any Kimi/Codex homes contain data you do not want this skill to read.
If your history may include secrets, try the skill first in a disposable profile or empty account.

Rated 2026-10-19 by an automated review of 25 files from the skill's repository.

Source

Repository: daymade/claude-code-skills (all skills from this repository)
Path: daymade-claude-code/read-claude-code-history/SKILL.md
Branch: main
Updated: 2026-09-20

read-codex-history — Reads, searches, and exports local OpenAI Codex history without continuing the old task.
recall — Reconstruct your recent working context from your own chat history, live state, and the shared record (user reports, prior fixes, incidents).
local-conversation-history — Entry point for local AI conversation history across providers.
audit-prep-assistant — Prepares codebases for security review using Trail of Bits' checklist.
automate-me — Use for "automate me", "create/update/refresh my -mode skill", "turn/capture my preferences or working style into a skill".
competitors — When the user wants to create competitor comparison or alternative pages for SEO and sales enablement.