thaitype/skills · Archived

claude-session-log

Show a full action log (text, tool calls, tool results) for a Claude Code session. Use when the user wants to replay, debug, or inspect what happened in a session, especially after running claude -p.

First seen Aug 20, 2026

Installation

$ npx skills add thaitype/skills --skill claude-session-log

Stronger alternatives

This repository is archived — consider an actively maintained alternative.

Similar popular skills

Related neighbors and high-traction skills in the same topics — useful to compare before installing.

Also in this package

Other skills from thaitype/skills · top by installs.

npx skills add thaitype/skills

Browse all from thaitype/skills

More details

Agent compatibility

Declared targets from SKILL.md / docs. Unmarked agents are not listed — the skill may still install via the CLI.

Claude Code Declared
Cursor Not declared
Codex Not declared
GitHub Copilot Not declared
Windsurf Not declared
Gemini CLI Not declared
Cline Not declared
OpenCode Not declared

Repository health

Stars 13
Default branch main
Open issues 0
Status Archived

Skill metadata

Parsed from SKILL.md frontmatter.

Declared agents claude-code

Package contents

Files included with this skill beyond the listing page.

  • skill md SKILL.md 2,456 B
  • docs SUMMARY.md 225 B

History

  1. First seen on skills.sh
  2. First recorded snapshot · 2 installs

SKILL.md

Claude Session Log

Replay a Claude Code session as a human-readable transcript or structured JSON. Shows assistant text, tool calls, and tool results for every turn.

Commands

# Show log for a specific session
python3 .claude/skills/claude-session-log/scripts/session_log.py <session-id>

# Show log for the most recent session
python3 .claude/skills/claude-session-log/scripts/session_log.py --latest

# JSON output (always full content, no truncation)
python3 .claude/skills/claude-session-log/scripts/session_log.py --latest --json

# Show only tool calls and results (skip pure text)
python3 .claude/skills/claude-session-log/scripts/session_log.py --latest --tools-only

# Disable truncation on tool results (default: 500 chars)
python3 .claude/skills/claude-session-log/scripts/session_log.py --latest --full

# Use custom projects directory
python3 .claude/skills/claude-session-log/scripts/session_log.py --latest --dir /path/to/projects

Output

Human-readable (default):

Session: 82c946c8-1eb8-4c70-9414-813fc0a278e4
Date: 2026-03-28 14:58
Model: claude-opus-4-6
Turns: 5

--- user (14:58:04) ---
Fix the bug in parser.py

--- assistant (14:58:06) ---
Let me read the file first.

  > Read scripts/parser.py
    [2847 chars, truncated]
    def parse(input):
        tokens = input.split()
        ...

  > Edit scripts/parser.py
    old: if t.startswith("--"):
    new: if t.startswith("--") or t.startswith("-"):

Fixed the parser to handle single-dash flags.

--- user (14:58:12) ---
  < tool_result (Read): [ok]
  < tool_result (Edit): [ok]

--- assistant (14:58:15) ---
Done. The parser now handles both --flag and -f syntax.

JSON mode (--json):

{
  "session_id": "82c946c8-...",
  "model": "claude-opus-4-6",
  "turns": 5,
  "messages": [
    {
      "role": "user",
      "timestamp": "2026-03-28T14:58:04Z",
      "content": [{"type": "text", "text": "Fix the bug..."}]
    },
    {
      "role": "assistant",
      "timestamp": "2026-03-28T14:58:06Z",
      "content": [
        {"type": "text", "text": "Let me read the file first."},
        {"type": "tool_use", "name": "Read", "input": {"file_path": "scripts/parser.py"}}
      ]
    }
  ]
}