my-project-public

repository

loading code, commits, and activity

repositories

loading repo index

#1	---
#2	name: session-logs
#3	description: Search and analyze your own session logs (older/parent conversations) using jq.
#4	metadata: {"clawdbot":{"emoji":"📜","requires":{"bins":["jq","rg"]}}}
#5	---
#6
#7	# session-logs
#8
#9	Search your complete conversation history stored in session JSONL files. Use this when a user references older/parent conversations or asks what was said before.
#10
#11	## Trigger
#12
#13	Use this skill when the user asks about prior chats, parent conversations, or historical context that isn’t in memory files.
#14
#15	## Location
#16
#17	Session logs live at: `~/.clawdbot/agents/<agentId>/sessions/` (use the `agent=<id>` value from the system prompt Runtime line).
#18
#19	- `sessions.json` - Index mapping session keys to session IDs
#20	- `<session-id>.jsonl` - Full conversation transcript per session
#21
#22	## Structure
#23
#24	Each `.jsonl` file contains messages with:
#25	- `type`: "session" (metadata) or "message"
#26	- `timestamp`: ISO timestamp
#27	- `message.role`: "user", "assistant", or "toolResult"
#28	- `message.content[]`: Text, thinking, or tool calls (filter `type=="text"` for human-readable content)
#29	- `message.usage.cost.total`: Cost per response
#30
#31	## Common Queries
#32
#33	### List all sessions by date and size
#34	```bash
#35	for f in ~/.clawdbot/agents/<agentId>/sessions/*.jsonl; do
#36	date=$(head -1 "$f" \| jq -r '.timestamp' \| cut -dT -f1)
#37	size=$(ls -lh "$f" \| awk '{print $5}')
#38	echo "$date $size $(basename $f)"
#39	done \| sort -r
#40	```
#41
#42	### Find sessions from a specific day
#43	```bash
#44	for f in ~/.clawdbot/agents/<agentId>/sessions/*.jsonl; do
#45	head -1 "$f" \| jq -r '.timestamp' \| grep -q "2026-01-06" && echo "$f"
#46	done
#47	```
#48
#49	### Extract user messages from a session
#50	```bash
#51	jq -r 'select(.message.role == "user") \| .message.content[]? \| select(.type == "text") \| .text' <session>.jsonl
#52	```
#53
#54	### Search for keyword in assistant responses
#55	```bash
#56	jq -r 'select(.message.role == "assistant") \| .message.content[]? \| select(.type == "text") \| .text' <session>.jsonl \| rg -i "keyword"
#57	```
#58
#59	### Get total cost for a session
#60	```bash
#61	jq -s '[.[] \| .message.usage.cost.total // 0] \| add' <session>.jsonl
#62	```
#63
#64	### Daily cost summary
#65	```bash
#66	for f in ~/.clawdbot/agents/<agentId>/sessions/*.jsonl; do
#67	date=$(head -1 "$f" \| jq -r '.timestamp' \| cut -dT -f1)
#68	cost=$(jq -s '[.[] \| .message.usage.cost.total // 0] \| add' "$f")
#69	echo "$date $cost"
#70	done \| awk '{a[$1]+=$2} END {for(d in a) print d, "$"a[d]}' \| sort -r
#71	```
#72
#73	### Count messages and tokens in a session
#74	```bash
#75	jq -s '{
#76	messages: length,
#77	user: [.[] \| select(.message.role == "user")] \| length,
#78	assistant: [.[] \| select(.message.role == "assistant")] \| length,
#79	first: .[0].timestamp,
#80	last: .[-1].timestamp
#81	}' <session>.jsonl
#82	```
#83
#84	### Tool usage breakdown
#85	```bash
#86	jq -r '.message.content[]? \| select(.type == "toolCall") \| .name' <session>.jsonl \| sort \| uniq -c \| sort -rn
#87	```
#88
#89	### Search across ALL sessions for a phrase
#90	```bash
#91	rg -l "phrase" ~/.clawdbot/agents/<agentId>/sessions/*.jsonl
#92	```
#93
#94	## Tips
#95
#96	- Sessions are append-only JSONL (one JSON object per line)
#97	- Large sessions can be several MB - use `head`/`tail` for sampling
#98	- The `sessions.json` index maps chat providers (discord, whatsapp, etc.) to session IDs
#99	- Deleted sessions have `.deleted.<timestamp>` suffix
#100
#101	## Fast text-only hint (low noise)
#102
#103	```bash
#104	jq -r 'select(.type=="message") \| .message.content[]? \| select(.type=="text") \| .text' ~/.clawdbot/agents/<agentId>/sessions/<id>.jsonl \| rg 'keyword'
#105	```
#106

z6Mkq5mY3JWtxoxUobWcfNHm7AkRubgSWEZTkBVqZXJviFZ5/my-project-public