CtrlK
BlogDocsLog inGet started
Tessl Logo

gitops-cluster-debug

Debug and troubleshoot Flux CD on live Kubernetes clusters (not local repo files) via the Flux MCP server — inspects Flux resource status, reads controller logs, traces dependency chains, and performs installation health checks. Use when users report failing, stuck, or not-ready Flux resources on a cluster, reconciliation errors, controller issues, artifact pull failures, image automation not updating tags, alerts or webhooks not being delivered, or need live cluster Flux Operator troubleshooting.

73

Quality

Does it follow best practices?

Impact

No eval scenarios have been run

SecuritybySnyk

Advisory

Suggest reviewing before use

SKILL.md
Quality
Evals
Security

Quality

Content

85%

Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.

A well-structured, highly actionable debugging skill with clear workflows, real bundle references, and explicit safety guards on the only destructive path. The main weakness is structural redundancy across the parallel HelmRelease/Kustomization/ResourceSet/Source workflows.

Suggestions

Factor the repeated per-workflow preamble (call get_flux_instance for controller status + apiVersion, then get_kubernetes_resources to analyze spec/status/inventory/events) into a shared 'Resource debugging template' referenced by Workflows 2-5 to reduce token cost.

Add an explicit validate/retry feedback loop for the apply path (e.g. apply manifest -> verify Ready condition -> if not Ready, inspect events and retry) rather than only the 'do not apply unless requested / overwrite=true' guard.

Collapse the duplicated step-10 RCA wording across Workflows 2, 3, and 5 into a single referenced 'Root cause analysis' template.

DimensionReasoningScore

Conciseness

Direct imperatives with no concept-padding (no explaining what Flux/Kubernetes is), but Workflows 2-5 repeat near-identical boilerplate ('Call get_flux_instance ... Call get_kubernetes_resources ... analyze spec, status, inventory, and events') that could be factored out and tightened.

2 / 3

Actionability

Names exact MCP tools and exact fields/conditions to inspect (e.g. 'chartRef or sourceRef', '.spec.git.push.branch', '$imagepolicy markers', 'Ready/FetchFailed/ArtifactInStorage'), giving copy-paste-ready diagnostic procedure without resorting to pseudocode.

3 / 3

Workflow Clarity

Eight clearly numbered, sequenced workflows with explicit cross-workflow routing (ResourceSet -> Workflow 2/3, Notification -> Workflow 8) and an explicit destructive-operation guard ('Do not apply resources unless explicitly requested'; 'errors unless overwrite is set to true').

3 / 3

Progressive Disclosure

SKILL.md serves as an overview with a CRD reference table linking to 18 verified schema files in assets/schemas/ and a 'Loading References' section pointing to two real one-level-deep reference files (flux-crds.md, troubleshooting.md), each with an explicit use-case signal.

3 / 3

Total

11

/

12

Passed

Description

100%

Based on the skill's description, can an agent find and select it at the right time? Clear, specific descriptions lead to better discovery.

A strong, third-person description that pairs concrete capabilities with an explicit, comprehensive 'Use when' trigger list and a clear disambiguating scope. No vague fluff or over-claims.

DimensionReasoningScore

Specificity

Lists multiple specific concrete actions — 'inspects Flux resource status, reads controller logs, traces dependency chains, and performs installation health checks' — rather than vague language.

3 / 3

Completeness

Explicitly states both what it does (debug/troubleshoot Flux via the MCP server with four named actions) and when to use it via an explicit 'Use when users report ...' clause.

3 / 3

Trigger Term Quality

Covers natural user-reported phrasings such as 'failing, stuck, or not-ready Flux resources', 'reconciliation errors', 'artifact pull failures', and 'alerts or webhooks not being delivered'.

3 / 3

Distinctiveness Conflict Risk

Narrow niche — live Flux CD on Kubernetes via flux-operator-mcp, explicitly 'not local repo files' — with distinct triggers unlikely to fire for unrelated skills.

3 / 3

Total

12

/

12

Passed

Validation

87%

Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.

Validation14 / 16 Passed

Validation for skill structure

CriteriaDescriptionResult

relative_links

Relative link issues: 19 deeper-than-1-level

Warning

referenced_paths_exist

Referenced path issues: 19 deeper-than-1-level

Warning

Total

14

/

16

Passed

Repository
fluxcd/agent-skills
Reviewed

Table of Contents

Is this your skill?

If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.