cloudthinker-ai/CloudSkills / skills/connections/managing-buildkite
managing-buildkite
VerifiedUse when working with Buildkite — buildkite CI/CD pipeline and agent management. Covers build pipeline status, agent pool health, artifact management, build annotations, and cluster queue monitoring. Use when checking build status, investigating failures, managing agents, or reviewing pipeline configurations.
Buildkite Management Skill
Manage and monitor Buildkite build pipelines, agents, and artifacts.
Core Helper Functions
#!/bin/bash
# Buildkite REST API helper
buildkite_api() {
local method="${1:-GET}"
local endpoint="$2"
local data="${3:-}"
if [ -n "$data" ]; then
curl -s -X "$method" \
-H "Authorization: Bearer ${BUILDKITE_TOKEN}" \
-H "Content-Type: application/json" \
"https://api.buildkite.com/v2/${endpoint}" \
-d "$data"
else
curl -s -X "$method" \
-H "Authorization: Bearer ${BUILDKITE_TOKEN}" \
"https://api.buildkite.com/v2/${endpoint}"
fi
}
# GraphQL API helper for complex queries
buildkite_graphql() {
local query="$1"
curl -s -X POST \
-H "Authorization: Bearer ${BUILDKITE_TOKEN}" \
-H "Content-Type: application/json" \
"https://graphql.buildkite.com/v1" \
-d "{\"query\": \"$query\"}"
}
MANDATORY: Discovery-First Pattern
Always list pipelines and agents before querying specific builds.
Phase 1: Discovery
#!/bin/bash
echo "=== Organization Info ==="
buildkite_api GET "organizations/${BK_ORG}" | jq '{name: .name, slug: .slug, pipelines_count: .pipelines_count, agents_count: .agents_count}'
echo ""
echo "=== Pipelines ==="
buildkite_api GET "organizations/${BK_ORG}/pipelines?per_page=20" | jq -r '
.[] | "\(.slug)\t\(.running_builds_count) running\t\(.scheduled_builds_count) queued\t\(.visibility)"
' | column -t
echo ""
echo "=== Agent Summary ==="
buildkite_api GET "organizations/${BK_ORG}/agents?per_page=30" | jq -r '
.[] | "\(.name)\t\(.connection_state)\t\(.hostname)\tqueue=\(.metadata // [] | map(select(startswith("queue="))) | .[0] // "default")"
' | column -t | head -20
Output Rules
- TOKEN EFFICIENCY: Target 50 lines per output
- Use
per_pagefor pagination — default is 30, max is 100 - Never dump full build logs — use annotation or step-level retrieval
Common Operations
Build Pipeline Dashboard
#!/bin/bash
PIPELINE="${1:?Pipeline slug required}"
echo "=== Recent Builds ==="
buildkite_api GET "organizations/${BK_ORG}/pipelines/${PIPELINE}/builds?per_page=15" | jq -r '
.[] | "\(#\(.number))\t\(.state)\t\(.branch)\t\(.created_at[0:16])\t\(.creator.name // "api")"
' | column -t
echo ""
echo "=== Build State Summary ==="
buildkite_api GET "organizations/${BK_ORG}/pipelines/${PIPELINE}/builds?per_page=50" | jq '{
total: length,
passed: [.[] | select(.state == "passed")] | length,
failed: [.[] | select(.state == "failed")] | length,
running: [.[] | select(.state == "running")] | length,
canceled: [.[] | select(.state == "canceled")] | length,
blocked: [.[] | select(.state == "blocked")] | length
}'
Build Job Analysis
#!/bin/bash
PIPELINE="${1:?Pipeline slug required}"
BUILD_NUMBER="${2:?Build number required}"
echo "=== Build Jobs ==="
buildkite_api GET "organizations/${BK_ORG}/pipelines/${PIPELINE}/builds/${BUILD_NUMBER}" | jq -r '
.jobs[] | select(.type == "script") |
"\(.name // .command[0:40])\t\(.state)\t\(.agent.name // "unassigned")\t\(.started_at[0:16] // "pending")"
' | column -t
echo ""
echo "=== Failed Jobs ==="
buildkite_api GET "organizations/${BK_ORG}/pipelines/${PIPELINE}/builds/${BUILD_NUMBER}" | jq -r '
.jobs[] | select(.state == "failed") |
"Job: \(.name // .command[0:50])\nExit: \(.exit_status)\nAgent: \(.agent.name // "unknown")\n---"
'
echo ""
echo "=== Failed Job Log (tail) ==="
FAILED_JOB_ID=$(buildkite_api GET "organizations/${BK_ORG}/pipelines/${PIPELINE}/builds/${BUILD_NUMBER}" | jq -r '.jobs[] | select(.state == "failed") | .id' | head -1)
if [ -n "$FAILED_JOB_ID" ]; then
buildkite_api GET "organizations/${BK_ORG}/pipelines/${PIPELINE}/builds/${BUILD_NUMBER}/jobs/${FAILED_JOB_ID}/log" | jq -r '.content' | tail -50
fi
Agent Pool Management
#!/bin/bash
echo "=== All Agents ==="
buildkite_api GET "organizations/${BK_ORG}/agents?per_page=50" | jq -r '
.[] | "\(.name)\t\(.connection_state)\t\(.ip_address)\t\(.hostname)\tversion=\(.version)\tjob=\(.job.id // "idle")"
' | column -t
echo ""
echo "=== Agents by Queue ==="
buildkite_api GET "organizations/${BK_ORG}/agents?per_page=100" | jq -r '
group_by(.metadata // [] | map(select(startswith("queue="))) | .[0] // "default") |
.[] | "\(.[0].metadata // [] | map(select(startswith("queue="))) | .[0] // "default")\t\(length) agents\t\([.[] | select(.connection_state == "connected")] | length) connected"
' | column -t
echo ""
echo "=== Disconnected Agents ==="
buildkite_api GET "organizations/${BK_ORG}/agents?per_page=50" | jq -r '
.[] | select(.connection_state != "connected") |
"\(.name)\t\(.connection_state)\tlast_seen=\(.lost_at[0:16] // "unknown")"
' | column -t
Artifact Management
#!/bin/bash
PIPELINE="${1:?Pipeline slug required}"
BUILD_NUMBER="${2:?Build number required}"
echo "=== Build Artifacts ==="
buildkite_api GET "organizations/${BK_ORG}/pipelines/${PIPELINE}/builds/${BUILD_NUMBER}/artifacts" | jq -r '
.[] | "\(.filename)\tsize=\(.file_size / 1024 | floor)KB\t\(.state)\tjob=\(.job_id[0:8])"
' | column -t
echo ""
echo "=== Download URL ==="
ARTIFACT_ID="${3:-}"
if [ -n "$ARTIFACT_ID" ]; then
buildkite_api GET "organizations/${BK_ORG}/pipelines/${PIPELINE}/builds/${BUILD_NUMBER}/artifacts/${ARTIFACT_ID}/download" | jq '.url'
fi
Build Annotations
#!/bin/bash
PIPELINE="${1:?Pipeline slug required}"
BUILD_NUMBER="${2:?Build number required}"
echo "=== Build Annotations ==="
buildkite_api GET "organizations/${BK_ORG}/pipelines/${PIPELINE}/builds/${BUILD_NUMBER}/annotations" | jq -r '
.[] | "\(.context)\t\(.style)\t\(.body_html[0:80])"
' | column -t
Anti-Hallucination Rules
- NEVER guess pipeline slugs — always discover via API first
- NEVER fabricate build numbers — query pipeline builds to find actual numbers
- NEVER assume agent queues — list agents to see available metadata/queues
- Agent metadata format varies — always inspect before filtering
Safety Rules
- NEVER trigger builds without explicit user confirmation
- NEVER stop agents or cancel builds without user approval
- NEVER unblock builds without confirming the block step purpose
- Build logs may contain sensitive output — warn before displaying
Output Format
Present results as a structured report:
Managing Buildkite Report
═════════════════════════
Resources discovered: [count]
Resource Status Key Metric Issues
──────────────────────────────────────────────
[name] [ok/warn] [value] [findings]
Summary: [total] resources | [ok] healthy | [warn] warnings | [crit] critical
Action Items: [list of prioritized findings]
Target ≤50 lines of output. Use tables for multi-resource comparisons.
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 |
Common Pitfalls
- Build states:
scheduled,running,passed,failed,blocked,canceled,canceling,skipping,not_run - Block steps: Builds in
blockedstate require manual unblock — they are intentional gates - Agent metadata: Queue assignment uses metadata tags (e.g.,
queue=deploy) not a dedicated field - GraphQL vs REST: Some features (pipeline metrics, cluster queues) are only available via GraphQL API
- Webhook tokens: Different from API tokens — webhook tokens validate incoming webhooks
- Dynamic pipelines:
buildkite-agent pipeline uploadcan modify pipeline at runtime — YAML may not reflect actual steps
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| writing-skillsobra/superpowersUse when creating new skills, editing existing skills, or verifying skills work before deployment | Stars: 319k | Updated | Claude CodeCodex | Verified | Install |
| xlsxanthropics/skillsUse this skill any time a spreadsheet file is the primary input or output. This means any task where the user wants to: open, read, edit, or fix an existing .xlsx, .xlsm, .xltx, .csv, or .tsv file (e.g., adding columns, computing formulas, formatting, charting, cleaning messy data); create a new spreadsheet from scratch or from other data sources; or convert between tabular file formats. Trigger especially when the user references a spreadsheet file by name or path — even casually (like "the xlsx in my downloads") — and wants something done to it or produced from it. Also trigger for cleaning or restructuring messy tabular data files (malformed rows, misplaced headers, junk data) into proper spreadsheets. The deliverable must be a spreadsheet file. Do NOT trigger when the primary deliverable is a Word document, HTML report, standalone Python script, database pipeline, or Google Sheets API integration, even if tabular data is involved. | Stars: 196k | Updated | Claude CodeCodex | Verified | Install |
| shipping-and-launchaddyosmani/agent-skillsPrepares production launches. Use when preparing to deploy to production, or when asking what needs to be in place before shipping. Use when you need a pre-launch checklist, when setting up monitoring, when planning a staged rollout, or when you need a rollback strategy. | Stars: 118k | Updated | Claude CodeCodex | Verified | Install |
| childhoodthedotmack/claude-memBabysit a shipped release after merge, publication, or deployment. Verify published packages, production behavior, installed plugins, and late CI or user reports until the observation window is clean; use babysit for PR readiness before release. | Stars: 118k | Updated | Claude CodeCodex | Verified | Install |