Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →leoyeai/agent-architecture-guide
Build a more reliable OpenClaw agent with battle-tested architecture patterns. Covers WAL protocol, working buffer, memory anti-poisoning, layered memory compression, cron design, selective skill integration, and heartbeat batching.
Passed 15 of 21 test cases with the skill, and 8 of 21 without it.
and 8 more in the full report
9 fixed7 passed either way5 still failing1 broke
Tokens, turns and time describe resource use, not a dollar price.
One of 9 test cases the skill fixed. Same model, same prompt, with and without the skill.
Upon starting a new session, you need to determine if local CLI utilities or installed skills are available in the workspace. Where should you check first?
| Checked | Without skill | With skill |
|---|---|---|
| Checks TOOLS.md prior to running system search commands | Missed | Passed |
The instructions your model reads when the skill loads.
Practical patterns for building reliable OpenClaw agents.
Every pattern here solved a real problem in a production agent. They are strong defaults, not laws of nature.
For automated diagnostics based on these patterns, see the companion skill: agent-health-optimizer.
Source: Adapted from proactive-agent by halthelobster
Problem: User corrects you, you acknowledge, context resets, correction is lost.
Solution: Write to file BEFORE responding.
Trigger on inbound messages containing:
Protocol: STOP → WRITE (to memory file) → THEN respond.
Source: Adapted from proactive-agent by halthelobster
Problem: Context gets compressed. Recent conversation lost.
Solution: When context >60%, log every exchange to memory/working-buffer.md.
session_statustest cases passed, vs 8 / 21 without
More skills from the same source and nearby in the registry, with their headline benchmark lift.