You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
The sampled observability coverage is healthy across the capped analysis set of 19 runs from the last 7 days, spanning 18 workflows. Every analyzed firewall-enabled run exposed an access.log, and every MCP-enabled run exposed MCP telemetry through rpc-messages.jsonl. No run in the sample was missing the core logs required to debug firewall egress or MCP activity.
The main gap is quality and richness rather than coverage. None of the sampled runs emitted preferred gateway.jsonl telemetry, so MCP analysis depended entirely on the canonical rpc-message/v2 fallback. That fallback was present and parseable in all 19 MCP-enabled runs, but it does not provide first-class per-call duration/status metrics the way structured gateway logs would.
Key Alerts and Anomalies
No critical issues detected.
Note
Core coverage is complete in this sample: firewall access.log coverage is 100.0% (19/19) and MCP telemetry coverage is 100.0% (19/19).
Warning
Preferred MCP gateway telemetry was absent in all 19 analyzed runs: gateway.jsonl coverage is 0/19, so duration and status metrics remain limited to what can be inferred from rpc-messages.jsonl.
Warning
Additional telemetry is slightly less complete than the core logs: safeoutputs.jsonl is missing in 3/19 analyzed runs.
Path discovery note: all sampled firewall logs were discovered recursively under each run directory; the sampled files resolved to the legacy path form sandbox/firewall/logs/access.log rather than squid-logs/access.log.
MCP Log Quality
Telemetry source: rpc-messages.jsonl fallback in all 19 MCP-enabled runs; gateway.jsonl was not present in the sample
Average response time: N/A from fallback schema without request/response duration pairing
Schema note: sampled MCP logs were valid rpc-message/v2 JSONL with event, direction, server_id, and payload; requests also carried top-level method.
Additional Telemetry
agent-stdio.log: 19/19
agent_output.json or agent_output/: 19/19
awf-config.json: 19/19
aw_info.json: 0/19; current sampled layout used awf-config.json instead
safeoutputs.jsonl: 16/19
Daily BYOK Ollama Test §31750168282 is missing safeoutputs.jsonl.
Daily AWF Spec Compiler Surfacing Review §31749758934 is missing safeoutputs.jsonl.
Daily Caveman Optimizer §31743356276 is missing safeoutputs.jsonl.
Healthy Runs Summary
All 19 analyzed runs had the minimum firewall and MCP artifacts needed for post-incident debugging.
Recommended Actions
Emit gateway.jsonl alongside rpc-messages.jsonl so per-call durations and statuses are available without reconstructing protocol traffic.
Keep the current firewall artifact export path under regression watch; this sample had 100% access.log coverage and should remain a hard requirement.
Investigate the missing safeoutputs.jsonl artifacts in the three runs noted above so secondary output auditing is as reliable as the core firewall and MCP logs.
📊 Historical Trends
No prior local baseline was available in this workspace, so this report is a point-in-time sample rather than a trend comparison.
reacted with thumbs up emoji reacted with thumbs down emoji reacted with laugh emoji reacted with hooray emoji reacted with confused emoji reacted with heart emoji reacted with rocket emoji reacted with eyes emoji
Uh oh!
There was an error while loading. Please reload this page.
Executive Summary
The sampled observability coverage is healthy across the capped analysis set of 19 runs from the last 7 days, spanning 18 workflows. Every analyzed firewall-enabled run exposed an
access.log, and every MCP-enabled run exposed MCP telemetry throughrpc-messages.jsonl. No run in the sample was missing the core logs required to debug firewall egress or MCP activity.The main gap is quality and richness rather than coverage. None of the sampled runs emitted preferred
gateway.jsonltelemetry, so MCP analysis depended entirely on the canonicalrpc-message/v2fallback. That fallback was present and parseable in all 19 MCP-enabled runs, but it does not provide first-class per-call duration/status metrics the way structured gateway logs would.Key Alerts and Anomalies
No critical issues detected.
Note
Core coverage is complete in this sample: firewall
access.logcoverage is 100.0% (19/19) and MCP telemetry coverage is 100.0% (19/19).Warning
Preferred MCP gateway telemetry was absent in all 19 analyzed runs:
gateway.jsonlcoverage is 0/19, so duration and status metrics remain limited to what can be inferred fromrpc-messages.jsonl.Warning
Additional telemetry is slightly less complete than the core logs:
safeoutputs.jsonlis missing in 3/19 analyzed runs.Coverage Summary
access.log)gateway.jsonlorrpc-messages.jsonl)📋 Detailed Run Analysis
Firewall-Enabled Runs
Missing Firewall Logs (access.log)
None.
MCP-Enabled Runs
Missing MCP Telemetry (no gateway.jsonl or rpc-messages.jsonl)
None.
🔍 Telemetry Quality Analysis
Firewall Log Quality
access.logentries analyzed: 4097sandbox/firewall/logs/access.lograther thansquid-logs/access.log.MCP Log Quality
rpc-messages.jsonlfallback in all 19 MCP-enabled runs;gateway.jsonlwas not present in the sampletools/callrequests: 139rpc-message/v2JSONL withevent,direction,server_id, andpayload; requests also carried top-levelmethod.Additional Telemetry
agent-stdio.log: 19/19agent_output.jsonoragent_output/: 19/19awf-config.json: 19/19aw_info.json: 0/19; current sampled layout usedawf-config.jsoninsteadsafeoutputs.jsonl: 16/19safeoutputs.jsonl.safeoutputs.jsonl.safeoutputs.jsonl.Healthy Runs Summary
All 19 analyzed runs had the minimum firewall and MCP artifacts needed for post-incident debugging.
Recommended Actions
gateway.jsonlalongsiderpc-messages.jsonlso per-call durations and statuses are available without reconstructing protocol traffic.access.logcoverage and should remain a hard requirement.safeoutputs.jsonlartifacts in the three runs noted above so secondary output auditing is as reliable as the core firewall and MCP logs.📊 Historical Trends
No prior local baseline was available in this workspace, so this report is a point-in-time sample rather than a trend comparison.
References: §31750168282, §31745489837, §31743356276
All reactions