hermes-backup/daily/2026-04-28_203212/skills/devops/arif-os-readiness-debugging/SKILL.md
arifOS endpoint/container debugging protocol — runtime witness before infrastructure speculation
npx skillsauth add ariffazil/openclaw-workspace arif-os-readiness-debuggingInstall this skill globally with one command. Works with Claude Code, Cursor, and Windsurf.
3 of 9 scanners reported clean
Some scanners were skipped, did not run, or reported a non-clean status. Review each row below.
Never debug infrastructure until the constitutional readiness surface has been read.
Authority order (highest to lowest):
/ready — constitutional readiness (14 checks, authoritative)/status.json — federation status/health — service health1. curl -s https://mcp.arif-fazil.com/ready
→ If status: pass → readiness is NOT the issue
→ If status: partial → READ the failures list
2. If /ready names a failing check (e.g., forge_dry_run_check):
→ Treat that as primary until disproven
→ Inspect source code for that check's implementation
→ Do NOT assume Docker bind / networking
3. If /ready passes but public endpoint fails:
→ Inspect Caddyfile inside running container:
docker exec caddy cat /etc/caddy/Caddyfile
→ Check Caddy logs:
docker logs caddy --tail 20
→ Verify reverse_proxy directives point to correct container service name
4. Only after /ready is clean AND the failure still persists:
→ docker exec <container> cat /proc/net/tcp (decode hex IP:port)
→ Check Docker network membership:
docker inspect <container> --format '{{json .NetworkSettings.Networks}}'
→ docker exec caddy nc -vz <target-service> <port>
→ docker exec caddy curl -v --connect-timeout 5 http://<target>:PORT/endpoint
ss or netstat without checking if they're installed — these are often absent
→ Use cat /proc/net/tcp + Python decode insteadss isn't in minimal containers — probe first with which ss || which netstatimport socket
STATE = {0x01:'ESTABLISHED', 0x06:'TIME_WAIT', 0x08:'CLOSE_WAIT', 0x0A:'LISTEN', 0x0B:'LISTEN'}
with open('/proc/net/tcp') as f:
for i, line in enumerate(f):
if i == 0: continue
p = line.split()
if len(p) < 10: continue
lip, lport = p[1].split(':')
rport = int(lport, 16)
state = int(p[3], 16)
ip = socket.inet_ntoa(bytes.fromhex(lip)[::-1])
sn = STATE.get(state, hex(state))
if rport == 8080 or state == 0x0A:
print(f'{sn} {ip}:{rport}')
"I have not yet proven transport failure. First witness /ready. Second witness Caddy route. Third witness container listener. Only then decide."
# 1. Wait for healthy
sleep 8 && docker ps --filter name=arifosmcp --format "{{.Status}}"
# 2. Check /ready
curl -s https://mcp.arif-fazil.com/ready | python3 -c "
import json,sys; d=json.load(sys.stdin)
print(f'Status: {d[\"status\"]}')
[print(f' {k}: {v[\"verdict\"]}') for k,v in d['checks'].items()]
print(f'Failures: {d[\"failures\"]}')
"
# 3. Check public endpoint
curl -s https://mcp.arif-fazil.com/health | python3 -m json.tool | head -10
curl -s https://mcp.arif-fazil.com/status.json | python3 -m json.tool | head -10
development
Governed intelligence skill for AAA as the abstraction, attestation, and abduction control plane across arifOS, APEX, A-FORGE, GEOX, WEALTH, WELL, and the ariffazil profile repository. Use when the user asks to explain or design AAA, route agentic work, reduce chaos/entropy in an arifOS federation task, create AREP/task declarations, classify risk, plan multi-repo changes, review governance boundaries, or translate human intent into evidence-backed, authority-safe, recursively agentic workflows. Provides deterministic F1-F13 floor checking, bounded abduction, and FederationReceipt composition.
development
Check every skill’s “use when” and “do not use when” clauses for collisions, missing negatives, and vague verbs like “help,” “assist,” or “improve.” Load when linting, reviewing, or validating trigger boundaries.
development
Bootstrap, design, and package new skills. Load when capturing user intent for a new skill or drafting its initial instruction framework.
content-media
Diagnose which federation services are up, down, or drifting. Produce a prioritized remediation plan.