lamm-mit/scienceclaw-status

ScienceClaw: Agent Status

Inspect the memory, activity, and health of a ScienceClaw agent.

When to use

Use this skill when the user asks to:

• Check what a ScienceClaw agent has been working on
• See recent investigations or journal entries
• Review agent activity stats (topics explored, tools used, post count)
• Search an agent's memory for a specific topic
• Get a health check on the ScienceClaw installation

How to run

Use bash to invoke the memory CLI. SCIENCECLAW_DIR defaults to ~/scienceclaw or ~/.scienceclaw/install.

SCIENCECLAW_DIR="${SCIENCECLAW_DIR:-$HOME/scienceclaw}"
AGENT=$(python3 -c "import json,pathlib; p=pathlib.Path.home()/'.scienceclaw'/'agent_profile.json'; print(json.loads(p.read_text()).get('name','ScienceAgent'))" 2>/dev/null || echo "ScienceAgent")
cd "$SCIENCECLAW_DIR"
python3 memory_cli stats --agent "$AGENT"

Subcommands

Journal stats (default status view)

python3 memory_cli journal --agent "$AGENT"

Recent journal entries

python3 memory_cli journal --agent "$AGENT" --recent 10

Active investigations

python3 memory_cli investigations --agent "$AGENT" --active

All investigated topics

python3 memory_cli journal --agent "$AGENT" --topics

Search memory for a topic

python3 memory_cli graph --agent "$AGENT" --search "CRISPR"

Full stats summary

python3 memory_cli stats --agent "$AGENT"

Parameters

• --agent — agent name (auto-resolved from ~/.scienceclaw/agent_profile.json if not provided)
• --recent N — show last N journal entries
• --active — show only in-progress investigations
• --search TERM — search knowledge graph for a term
• --topics — list all topics ever investigated

Resolving the agent name

Always auto-resolve the agent name from the profile before running:

AGENT=$(python3 -c "
import json, pathlib
p = pathlib.Path.home() / '.scienceclaw' / 'agent_profile.json'
print(json.loads(p.read_text()).get('name', 'ScienceAgent'))
" 2>/dev/null || echo "ScienceAgent")

If the user specifies an agent name explicitly, use that instead.

Workspace context

If the workspace memory (memory.md) contains a preferred agent name, use that as the default --agent value.

After running

Report back to the user:

• Agent name and total journal entries
• Number of topics investigated
• Most recent 3–5 topics (if available)
• Any active/in-progress investigations
• Offer to run a deeper search with scienceclaw-investigate on any of the listed topics