Chat Log Search
Search local Codex, Claude Code, and Cursor chat logs with layered source and detail controls.
Scripts
| Script |
Scope |
scripts/chat_log_search.py |
All providers |
scripts/codex_log_search.py |
Codex only |
scripts/claude_log_search.py |
Claude Code only |
scripts/cursor_log_search.py |
Cursor only |
Source Layers (--source)
history — fast prompt index (~/.codex/history.jsonl, ~/.claude/history.jsonl)sessions — full session / transcript logsboth — history index plus session logs
Cursor has no history index; use --source sessions for Cursor.
Detail Layers (--detail)
prompts — user prompts onlymessages — user + assistant text, no tools or thinkingfinal — user prompts + final assistant reply per turn (no intermediate tool-step chatter)tools — tool inputs and outputs onlyfull — all extractable events (text, tools, reasoning, raw session events)
Quick Start
Fast user prompt search (all providers with history indexes):
python3 skills/mem/chat-log-search/scripts/chat_log_search.py "linear ticket" --source history --detail prompts --limit 20
Codex clean transcript (user + final assistant replies):
python3 skills/mem/chat-log-search/scripts/codex_log_search.py "memory system" --source sessions --detail final --limit 50
Claude tool audit:
python3 skills/mem/chat-log-search/scripts/claude_log_search.py "posthog" --source sessions --detail tools --limit 50
Cursor full session dump search:
python3 skills/mem/chat-log-search/scripts/cursor_log_search.py "paywall" --source sessions --detail full --limit 50
Everything across history + sessions:
python3 skills/mem/chat-log-search/scripts/chat_log_search.py "sentry" --source both --detail full --limit 50
JSON output:
python3 skills/mem/chat-log-search/scripts/claude_log_search.py "council" --source both --detail messages --json
Workflow
- Start with
--source history --detail prompts to find candidate sessions quickly.
- Re-run the matching session with
--source sessions --detail final for a readable transcript.
- Use
--detail tools when debugging agent behavior or MCP/tool usage.
- Use
--detail full when you need reasoning, raw events, or complete context.
- Narrow noisy results with
--after, --before, or --session-id.
- Add
--exclude-approval-review when mining user conversations or recurring patterns so Guardian assessment sessions do not duplicate injected transcripts.
Output
- Human-readable default:
timestamp | provider | source | detail | kind | role | session_id | snippet
- JSON records include:
provider, source, detail, kind, role, session_id, timestamp, text, file, meta
Recency & limits
- Default:
--limit 50 --order recent → 50 newest matching records
- All matches:
--all or --limit 0
- Scan depth:
--recent-window N → only search last N raw lines/records per file/session
- Turn-based limit:
--detail final --limit-units turns --limit N
Full option reference: references/search_api.md
Notes
- Read-only. Do not modify files under
~/.codex, ~/.claude, or ~/.cursor.
- Codex archived sessions are included by default; pass
--no-archived on Codex/all scripts.
- Codex/all-provider scripts support
--exclude-approval-review. It skips Guardian approval-review session logs from session metadata while preserving normal tasks, normal subagents, and history-index prompts.
- Cursor prefers
agent-transcripts; pass --include-store-db on Cursor/all scripts only if needed.
- Log locations and formats:
references/log_locations.md
- Detail layer reference:
references/search_layers.md
1---2name: grim-mem-chat-log-search-23description: (grim:mem:chat-log-search): (Chat Log Search): Search and summarize past Codex Desktop, Claude Code, and Cursor chat logs with layered detail (prompts, messages, final replies, tools, full). Use when asked to find prior conversations, prompts, decisions, tool calls, or to locate a session by keyword/date/session id.4---5# Chat Log Search67Search local Codex, Claude Code, and Cursor chat logs with layered source and detail controls.89## Scripts1011| Script | Scope |12|---|---|13| `scripts/chat_log_search.py` | All providers |14| `scripts/codex_log_search.py` | Codex only |15| `scripts/claude_log_search.py` | Claude Code only |16| `scripts/cursor_log_search.py` | Cursor only |1718## Source Layers (`--source`)1920- `history` — fast prompt index (`~/.codex/history.jsonl`, `~/.claude/history.jsonl`)21- `sessions` — full session / transcript logs22- `both` — history index plus session logs2324Cursor has no history index; use `--source sessions` for Cursor.2526## Detail Layers (`--detail`)2728- `prompts` — user prompts only29- `messages` — user + assistant text, no tools or thinking30- `final` — user prompts + final assistant reply per turn (no intermediate tool-step chatter)31- `tools` — tool inputs and outputs only32- `full` — all extractable events (text, tools, reasoning, raw session events)3334## Quick Start3536Fast user prompt search (all providers with history indexes):37```bash38python3 skills/mem/chat-log-search/scripts/chat_log_search.py "linear ticket" --source history --detail prompts --limit 2039```4041Codex clean transcript (user + final assistant replies):42```bash43python3 skills/mem/chat-log-search/scripts/codex_log_search.py "memory system" --source sessions --detail final --limit 5044```4546Claude tool audit:47```bash48python3 skills/mem/chat-log-search/scripts/claude_log_search.py "posthog" --source sessions --detail tools --limit 5049```5051Cursor full session dump search:52```bash53python3 skills/mem/chat-log-search/scripts/cursor_log_search.py "paywall" --source sessions --detail full --limit 5054```5556Everything across history + sessions:57```bash58python3 skills/mem/chat-log-search/scripts/chat_log_search.py "sentry" --source both --detail full --limit 5059```6061JSON output:62```bash63python3 skills/mem/chat-log-search/scripts/claude_log_search.py "council" --source both --detail messages --json64```6566## Workflow67681. Start with `--source history --detail prompts` to find candidate sessions quickly.692. Re-run the matching session with `--source sessions --detail final` for a readable transcript.703. Use `--detail tools` when debugging agent behavior or MCP/tool usage.714. Use `--detail full` when you need reasoning, raw events, or complete context.725. Narrow noisy results with `--after`, `--before`, or `--session-id`.736. Add `--exclude-approval-review` when mining user conversations or recurring patterns so Guardian assessment sessions do not duplicate injected transcripts.7475## Output7677- Human-readable default:78 `timestamp | provider | source | detail | kind | role | session_id | snippet`79- JSON records include: `provider`, `source`, `detail`, `kind`, `role`, `session_id`, `timestamp`, `text`, `file`, `meta`8081## Recency & limits8283- Default: `--limit 50 --order recent` → **50 newest matching records**84- All matches: `--all` or `--limit 0`85- Scan depth: `--recent-window N` → only search last N raw lines/records per file/session86- Turn-based limit: `--detail final --limit-units turns --limit N`8788Full option reference: `references/search_api.md`8990## Notes9192- Read-only. Do not modify files under `~/.codex`, `~/.claude`, or `~/.cursor`.93- Codex archived sessions are included by default; pass `--no-archived` on Codex/all scripts.94- Codex/all-provider scripts support `--exclude-approval-review`. It skips Guardian approval-review session logs from session metadata while preserving normal tasks, normal subagents, and history-index prompts.95- Cursor prefers `agent-transcripts`; pass `--include-store-db` on Cursor/all scripts only if needed.96- Log locations and formats: `references/log_locations.md`97- Detail layer reference: `references/search_layers.md`
MindGoblinStudios