Security checks across malware telemetry and agentic risk
This skill is mostly coherent, but it can make high-impact OpenClaw configuration changes from AI-generated proposals and run live post-apply checks with limited enforced user review.
Review the generated plan before applying, avoid putting secrets in requirements or proposals, expect gateway restarts and live agent canary messages during apply, and use a private trusted backup directory. The package metadata says this slug is deprecated in favor of oc-guard-skill, so prefer the maintained release if available.
ts = dt.datetime.now().strftime("%Y%m%d_%H%M%S")
backup_file = BACKUP_DIR / f"openclaw.json.{ts}.bak"
dump_json(LAST_PROPOSAL_PATH, proposal)
shutil.copy2(config_file, backup_file)
dump_json(config_file, modified)
log(f"Backup created: {backup_file}")"stdout": stdout,
"stderr": stderr,
}
OPENCODE_DEBUG_PATH.write_text(
json.dumps(payload, ensure_ascii=False, indent=2) + "\n",
encoding="utf-8",
)restarted = run([str(OPENCLAW_BIN), "gateway", "restart"], timeout=60, check=False)
if restarted.returncode != 0:
log("Gateway restart failed, rolling back")
shutil.copy2(backup_file, config_file)
run([str(OPENCLAW_BIN), "gateway", "restart"], timeout=60, check=False)
fail("restart failed after apply; rolled back")ok, status_text = check_gateway_running()
if not ok:
log("Gateway unhealthy after apply, rolling back")
shutil.copy2(backup_file, config_file)
run([str(OPENCLAW_BIN), "gateway", "restart"], timeout=60, check=False)
fail("gateway unhealthy after apply; rolled back")"extracted_json": text,
"parse_error": str(e),
}
OPENCODE_DEBUG_PATH.write_text(
json.dumps(payload, ensure_ascii=False, indent=2) + "\n",
encoding="utf-8",
)canary_notes = run_post_apply_canary(modified)
except Exception as e:
log(f"Post-apply canary failed, rolling back: {e}")
shutil.copy2(backup_file, config_file)
run([str(OPENCLAW_BIN), "gateway", "restart"], timeout=60, check=False)
fail(f"post-apply canary failed; rolled back: {e}")65/65 vendors flagged this skill as clean.