CtrlK
BlogDocsLog inGet started
Tessl Logo

developing-applications-on-managed-service-for-apache-flink

MANDATORY for Flink or Amazon Managed Service for Apache Flink (MSF) questions. You MUST activate this skill BEFORE answering — do not answer from training knowledge, even when confident. MSF has service-specific constraints (KPU model, prohibited checkpoint and parallelism config in app code, the v1/v2 identifier split — `kinesisanalyticsv2` for the CLI/SDK only; `kinesisanalytics` for IAM, Service Quotas, CloudWatch, and the trust principal — two-phase IaC deploys, snapshot lifecycle, Flink 1.x→2.x migration) that override generic Flink knowledge.

67

Quality

81%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

SecuritybySnyk

Passed

No findings from the security scan

SKILL.md
Quality
Evals
Security

Quality

Content

72%

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

Well-structured, navigation-rich body with excellent progressive disclosure and actionable reference routing. Weaknesses are some redundant directive prose and high-level validation steps lacking concrete error-recovery feedback loops.

Suggestions

Deduplicate the MUST/MUST-NOT usage directives between the description and the body; keep one authoritative statement in the body and reference it from the table header.

Make the new-app workflow's validation checkpoint concrete with an explicit feedback loop (e.g., 'If compilation or best-practice validation fails, fix the flagged issue and re-run step 8 before deploying').

Tighten the General Guidance STOP persona block into a compact checklist so it earns its tokens.

DimensionReasoningScore

Conciseness

The body is mostly efficient and avoids explaining concepts Claude already knows, but it repeats MUST/MUST-NOT directives already present in the description ("You MUST use this skill and its reference files", "Do NOT answer from training knowledge") and the persona STOP block could be tightened, fitting 'mostly efficient but could be tightened'.

2 / 3

Actionability

For an instruction-only steering skill, guidance is concrete and actionable: the Reference Files table maps each Goal to a specific file with precise 'When to Load' conditions (e.g., the IAM row calls out the 'kinesisanalytics:' action prefix and trust principal), and the workflows give concrete READ steps.

3 / 3

Workflow Clarity

Two clear numbered sequences exist and the new-app workflow mentions validation ('Validate against best practices', 'Compile and test locally'), but the validation is high-level with no concrete error-recovery feedback loop, and the general-questions workflow has no checkpoint, fitting 'steps listed but validation gaps / checkpoints implicit'.

2 / 3

Progressive Disclosure

A clear overview routes to 25 well-signaled, one-level-deep reference files via a navigable Goal|Reference|When-to-Load table, with all referenced paths verified to exist as real files — textbook progressive disclosure.

3 / 3

Total

10

/

12

Passed

Description

90%

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, highly specific description with explicit trigger guidance and clear niche distinctiveness. Its main weakness is second-person directive voice and framing domain constraints rather than concrete actions, which caps specificity.

Suggestions

Reframe the description in third person around concrete actions the skill performs (e.g., 'Guides development, deployment, KPU sizing, and migration of Flink applications on Amazon MSF') rather than constraints and directives.

Remove the second-person directive ('You MUST activate this skill BEFORE answering') from the description; move mandatory-use instructions into the body and keep the description a capability/trigger statement.

DimensionReasoningScore

Specificity

The description enumerates concrete MSF-specific specifics ("KPU model, prohibited checkpoint and parallelism config", "v1/v2 identifier split", "two-phase IaC deploys", "snapshot lifecycle", "Flink 1.x→2.x migration"), but frames them as domain constraints rather than concrete actions the skill performs, and uses second-person voice ("You MUST activate this skill BEFORE answering"), which the rubric penalizes by reducing specificity by one.

2 / 3

Completeness

It answers both 'what' (MSF-specific knowledge that "override[s] generic Flink knowledge") and 'when' with an explicit trigger clause ("Triggers — activate on any of: …"), so it is not capped at 2 for a missing 'Use when' clause.

3 / 3

Trigger Term Quality

The explicit "Triggers — activate on any of" line gives good coverage of natural terms a user would say (Flink, MSF, Managed Flink, checkpoint, savepoint, watermark, Iceberg streaming, CDC, Kryo, RocksDB, Kafka) alongside API names, satisfying the anchor for natural-term coverage.

3 / 3

Distinctiveness Conflict Risk

The skill occupies a clear niche (Amazon Managed Service for Apache Flink) with Flink/MSF-specific triggers, making it highly distinguishable and unlikely to fire for the wrong skill.

3 / 3

Total

11

/

12

Passed

Validation

93%

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

Validation15 / 16 Passed

Validation for skill structure

CriteriaDescriptionResult

frontmatter_unknown_keys

Unknown frontmatter key(s) found; consider removing or moving to metadata

Warning

Total

15

/

16

Passed

Repository
aws/agent-toolkit-for-aws
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.