Consul KV Management Skill

Monitor, analyze, and optimize Consul KV store operations safely.

MANDATORY: Discovery-First Pattern

Always check Consul agent status and list KV prefixes before any key operations. Never assume key paths or session states.

Phase 1: Discovery

#!/bin/bash

CONSUL_ADDR="${CONSUL_HTTP_ADDR:-http://localhost:8500}"
CONSUL_TOKEN="${CONSUL_HTTP_TOKEN:+X-Consul-Token: $CONSUL_HTTP_TOKEN}"

echo "=== Agent Self ==="
curl -s ${CONSUL_TOKEN:+-H "$CONSUL_TOKEN"} "$CONSUL_ADDR/v1/agent/self" | python3 -c "
import sys, json
data = json.load(sys.stdin)
config = data.get('Config', {})
print(f\"Datacenter: {config.get('Datacenter','?')}\")
print(f\"Node: {config.get('NodeName','?')}\")
print(f\"Server: {config.get('Server', False)}\")
print(f\"Version: {config.get('Version','?')}\")
" 2>/dev/null

echo ""
echo "=== Cluster Members ==="
curl -s ${CONSUL_TOKEN:+-H "$CONSUL_TOKEN"} "$CONSUL_ADDR/v1/agent/members" | python3 -c "
import sys, json
for m in json.load(sys.stdin):
    print(f\"  {m['Name']}: {m['Addr']}:{m['Port']} | Status={m['Status']} | Type={'server' if m.get('Tags',{}).get('role')=='consul' else 'client'}\")
" 2>/dev/null

echo ""
echo "=== KV Top-Level Keys ==="
curl -s ${CONSUL_TOKEN:+-H "$CONSUL_TOKEN"} "$CONSUL_ADDR/v1/kv/?keys&separator=/" | python3 -c "
import sys, json
keys = json.load(sys.stdin)
print(f\"Top-level prefixes: {len(keys)}\")
for k in keys[:20]:
    print(f\"  {k}\")
" 2>/dev/null

echo ""
echo "=== Active Sessions ==="
curl -s ${CONSUL_TOKEN:+-H "$CONSUL_TOKEN"} "$CONSUL_ADDR/v1/session/list" | python3 -c "
import sys, json
sessions = json.load(sys.stdin)
print(f\"Active sessions: {len(sessions)}\")
for s in sessions[:10]:
    print(f\"  {s['ID'][:12]}... | Node: {s.get('Node','?')} | TTL: {s.get('TTL','none')} | Behavior: {s.get('Behavior','?')}\")
" 2>/dev/null

Phase 1 outputs: Datacenter, node info, cluster members, KV namespace tree, active sessions.

Phase 2: Analysis

#!/bin/bash

CONSUL_ADDR="${CONSUL_HTTP_ADDR:-http://localhost:8500}"
CONSUL_TOKEN="${CONSUL_HTTP_TOKEN:+X-Consul-Token: $CONSUL_HTTP_TOKEN}"
PREFIX="${1:-}"

echo "=== Key Count by Prefix ==="
curl -s ${CONSUL_TOKEN:+-H "$CONSUL_TOKEN"} "$CONSUL_ADDR/v1/kv/$PREFIX?keys" | python3 -c "
import sys, json
keys = json.load(sys.stdin)
print(f\"Total keys under '{sys.argv[1] if len(sys.argv)>1 else '/'}': {len(keys)}\")
prefixes = {}
for k in keys:
    parts = k.strip('/').split('/')
    if parts:
        prefixes[parts[0]] = prefixes.get(parts[0], 0) + 1
for p, c in sorted(prefixes.items(), key=lambda x: -x[1])[:15]:
    print(f\"  {p}/: {c} keys\")
" "$PREFIX" 2>/dev/null

echo ""
echo "=== Locked Keys ==="
curl -s ${CONSUL_TOKEN:+-H "$CONSUL_TOKEN"} "$CONSUL_ADDR/v1/kv/$PREFIX?recurse" | python3 -c "
import sys, json
data = json.load(sys.stdin)
locked = [kv for kv in data if kv.get('Session')]
print(f\"Locked keys: {len(locked)}\")
for kv in locked[:10]:
    print(f\"  {kv['Key']}: session={kv['Session'][:12]}...\")
" 2>/dev/null || echo "No locked keys"

echo ""
echo "=== Prepared Queries ==="
curl -s ${CONSUL_TOKEN:+-H "$CONSUL_TOKEN"} "$CONSUL_ADDR/v1/query" | python3 -c "
import sys, json
queries = json.load(sys.stdin)
print(f\"Prepared queries: {len(queries)}\")
for q in queries[:5]:
    print(f\"  {q.get('Name','unnamed')}: service={q.get('Service',{}).get('Service','?')}\")
" 2>/dev/null || echo "No prepared queries"

Output Format

CONSUL KV ANALYSIS
==================
Datacenter: [dc] | Node: [name] | Members: [count]
KV Prefixes: [count] | Total Keys: [count] | Sessions: [count]

ISSUES FOUND:
- [issue with affected prefix/session]

RECOMMENDATIONS:
- [actionable recommendation]

Anti-Hallucination Rules

1. NEVER assume resource names — always discover via CLI/API in Phase 1 before referencing in Phase 2.
2. NEVER fabricate metric names or dimensions — verify against the service documentation or --help output.
3. NEVER mix CLI commands between service versions — confirm which version/API you are targeting.
4. ALWAYS use the discovery → verify → analyze chain — every resource referenced must have been discovered first.
5. ALWAYS handle empty results gracefully — an empty response is valid data, not an error to retry.

Counter-Rationalizations

Shortcut	Counter	Why
"I'll skip discovery and check known resources"	Always run Phase 1 discovery first	Resource names change, new resources appear — assumed names cause errors
"The user only asked for a quick check"	Follow the full discovery → analysis flow	Quick checks miss critical issues; structured analysis catches silent failures
"Default configuration is probably fine"	Audit configuration explicitly	Defaults often leave logging, security, and optimization features disabled
"Metrics aren't needed for this"	Always check relevant metrics when available	API/CLI responses show current state; metrics reveal trends and intermittent issues
"I don't have access to that"	Try the command and report the actual error	Assumed permission failures prevent useful investigation; actual errors are informative