cloudthinker-ai Tekton Management Skill
Manage and monitor Tekton Pipelines, Tasks, and Triggers on Kubernetes.
Core Helper Functions
#!/bin/bash
# Tekton CLI wrapper with namespace
tkn_cmd() {
tkn "$@" --namespace "${TEKTON_NAMESPACE:-default}" 2>/dev/null
}
# kubectl wrapper for Tekton CRDs
tkn_k8s() {
kubectl "$@" -n "${TEKTON_NAMESPACE:-default}" 2>/dev/null
}
MANDATORY: Discovery-First Pattern
Always list available pipelines, tasks, and triggers before querying specific runs.
Phase 1: Discovery
#!/bin/bash
NS="${TEKTON_NAMESPACE:-default}"
echo "=== Tekton Pipelines ==="
tkn_cmd pipeline list || \
tkn_k8s get pipelines -o custom-columns="NAME:.metadata.name,AGE:.metadata.creationTimestamp"
echo ""
echo "=== Tekton Tasks ==="
tkn_cmd task list || \
tkn_k8s get tasks -o custom-columns="NAME:.metadata.name,AGE:.metadata.creationTimestamp"
echo ""
echo "=== ClusterTasks ==="
tkn clustertask list 2>/dev/null || \
kubectl get clustertasks -o custom-columns="NAME:.metadata.name" 2>/dev/null
echo ""
echo "=== Recent PipelineRuns ==="
tkn_cmd pipelinerun list --limit 15 || \
tkn_k8s get pipelineruns -o custom-columns="NAME:.metadata.name,PIPELINE:.spec.pipelineRef.name,STATUS:.status.conditions[0].reason,START:.status.startTime" --sort-by=.status.startTime | tail -15
echo ""
echo "=== EventListeners ==="
tkn_k8s get eventlisteners -o custom-columns="NAME:.metadata.name,STATUS:.status.conditions[0].reason" 2>/dev/null
Output Rules
- TOKEN EFFICIENCY: Target 50 lines per output
- Use
tknCLI for human-friendly output,kubectl -o json | jqfor programmatic queries - Never dump full task logs — extract step-level output
Common Operations
PipelineRun Status Dashboard
#!/bin/bash
NS="${TEKTON_NAMESPACE:-default}"
echo "=== PipelineRun Summary ==="
tkn_k8s get pipelineruns -o json | jq '{
total: (.items | length),
succeeded: [.items[] | select(.status.conditions[0].reason == "Succeeded")] | length,
failed: [.items[] | select(.status.conditions[0].reason == "Failed")] | length,
running: [.items[] | select(.status.conditions[0].reason == "Running")] | length,
pending: [.items[] | select(.status.conditions[0].reason == "PipelineRunPending")] | length
}'
echo ""
echo "=== Failed PipelineRuns ==="
tkn_k8s get pipelineruns -o json | jq -r '
.items[] | select(.status.conditions[0].reason == "Failed") |
"\(.metadata.name)\t\(.spec.pipelineRef.name // "inline")\t\(.status.conditions[0].message[0:60])"
' | column -t | head -10
echo ""
echo "=== Running Pipelines ==="
tkn_cmd pipelinerun list --label tekton.dev/pipeline --limit 10 2>/dev/null || \
tkn_k8s get pipelineruns --field-selector="status.conditions[0].reason=Running" -o custom-columns="NAME:.metadata.name,PIPELINE:.spec.pipelineRef.name,START:.status.startTime"
TaskRun Analysis
#!/bin/bash
PIPELINE_RUN="${1:?PipelineRun name required}"
NS="${TEKTON_NAMESPACE:-default}"
echo "=== TaskRuns in PipelineRun ==="
tkn_cmd pipelinerun describe "$PIPELINE_RUN" 2>/dev/null || \
tkn_k8s get taskruns -l "tekton.dev/pipelineRun=${PIPELINE_RUN}" -o json | jq -r '
.items[] |
"\(.metadata.name)\t\(.metadata.labels["tekton.dev/pipelineTask"])\t\(.status.conditions[0].reason)\t\(.status.startTime[0:16])"
' | column -t
echo ""
echo "=== Failed TaskRun Logs ==="
FAILED_TR=$(tkn_k8s get taskruns -l "tekton.dev/pipelineRun=${PIPELINE_RUN}" -o json | jq -r '
.items[] | select(.status.conditions[0].reason == "Failed") | .metadata.name' | head -1)
if [ -n "$FAILED_TR" ]; then
tkn_cmd taskrun logs "$FAILED_TR" 2>/dev/null || \
tkn_k8s get taskrun "$FAILED_TR" -o json | jq -r '
.status.steps[] | select(.terminated.exitCode != 0) |
"Step: \(.name)\nExit: \(.terminated.exitCode)\nReason: \(.terminated.reason)"
'
fi
Task Management
#!/bin/bash
NS="${TEKTON_NAMESPACE:-default}"
echo "=== Pipeline Definition ==="
PIPELINE_NAME="${1:?Pipeline name required}"
tkn_k8s get pipeline "$PIPELINE_NAME" -o json | jq '{
name: .metadata.name,
params: [.spec.params[]? | {name, type, default}],
tasks: [.spec.tasks[] | {name, taskRef: .taskRef.name, runAfter: .runAfter}],
workspaces: [.spec.workspaces[]? | .name]
}'
echo ""
echo "=== Task Details ==="
TASK_NAME="${2:-}"
if [ -n "$TASK_NAME" ]; then
tkn_k8s get task "$TASK_NAME" -o json | jq '{
name: .metadata.name,
params: [.spec.params[]? | {name, type, default}],
steps: [.spec.steps[] | {name, image}],
workspaces: [.spec.workspaces[]? | .name]
}'
fi
Trigger Bindings & Templates
#!/bin/bash
NS="${TEKTON_NAMESPACE:-default}"
echo "=== TriggerBindings ==="
tkn_k8s get triggerbindings -o json | jq -r '
.items[] | "\(.metadata.name)\tparams=\([.spec.params[]? | .name] | join(","))"
' | column -t
echo ""
echo "=== TriggerTemplates ==="
tkn_k8s get triggertemplates -o json | jq -r '
.items[] | "\(.metadata.name)\tparams=\([.spec.params[]? | .name] | join(","))"
' | column -t
echo ""
echo "=== EventListeners ==="
tkn_k8s get eventlisteners -o json | jq -r '
.items[] |
"\(.metadata.name)\tstatus=\(.status.conditions[0].reason // "unknown")\taddress=\(.status.address.url // "pending")"
' | column -t
Pipeline Execution
#!/bin/bash
PIPELINE_NAME="${1:?Pipeline name required}"
DRY_RUN="${2:-true}"
if [ "$DRY_RUN" = "true" ]; then
echo "=== DRY RUN: Pipeline params and workspaces ==="
tkn_k8s get pipeline "$PIPELINE_NAME" -o json | jq '{
required_params: [.spec.params[]? | select(.default == null) | .name],
optional_params: [.spec.params[]? | select(.default != null) | {name, default}],
workspaces: [.spec.workspaces[]? | {name, optional: .optional}]
}'
echo ""
echo "To execute, use: tkn pipeline start $PIPELINE_NAME -p key=value -w name=pvc,claimName=my-pvc"
else
echo "=== Starting Pipeline ==="
tkn_cmd pipeline start "$PIPELINE_NAME" --showlog
fi
Anti-Hallucination Rules
- NEVER guess PipelineRun or TaskRun names — always list first
- NEVER fabricate task step names — inspect the task definition
- NEVER assume namespace — Tekton resources are namespace-scoped
- ClusterTasks are cluster-scoped — do not add namespace when querying
Safety Rules
- NEVER start pipelines without explicit user confirmation
- NEVER delete PipelineRuns or TaskRuns without user approval
- NEVER modify TriggerBindings without understanding downstream effects
- Pipeline execution may provision cloud resources — confirm intent
Output Format
Present results as a structured report:
Managing Tekton 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
- tkn vs kubectl:
tknCLI provides better UX but requires installation — fall back to kubectl - ClusterTask deprecation: ClusterTasks are deprecated in favor of Tekton Resolver — check version
- Workspace binding: PipelineRuns must bind workspaces — missing bindings cause failures
- Finally tasks:
finallytasks always run regardless of pipeline success — check for cleanup tasks - Result propagation: Task results are passed via
$(tasks.taskname.results.resultname)— verify names exactly - Timeout defaults: Default pipeline timeout is 1 hour — long builds may be killed silently