CtrlK
BlogDocsLog inGet started
Tessl Logo
SkillPublic

cloud-overview

Cloud exploitation lane — AWS IAM privesc, S3 takeover, k8s RBAC abuse, Terraform state leaks, cloud metadata pivoting.

Quality

76%

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

Fix and improve this skill with Tessl

tessl review fix ./packages/decepticon/decepticon/skills/standard/cloud/SKILL.md
SKILL.md
Quality
Evals
Security

Quality

Content

72%Weight 40%Scale 1-5

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

An extremely lean, well-structured routing catalog: a playbook table and two crisp workflows with real AWS/kubectl commands. The weaknesses are undefined internal tool invocations (no syntax or pointer for the *_audit / graph functions) and the absence of explicit validation checkpoints in the workflows.

Suggestions

Define or link the internal tool calls: give the invocation syntax or a pointer for iam_policy_audit, tfstate_audit, k8s_audit, kg_add_node, metadata_endpoints, and plan_attack_chains, or reference the playbook file that defines each.

Add validation checkpoints, e.g. verify the STS identity and assumed role before enumerating policies, and confirm extracted credentials are live before creating the credential node and 'leaks' edge.

Replace the "<me>" placeholder and 'fetch JSON' step with the exact command (e.g. aws iam get-policy-version) so every numbered step is copy-paste executable.

DimensionReasoningScore

Conciseness

The body is a lean catalog table plus two terse numbered workflows; there is no padding, no explanation of concepts Claude already knows, and every line carries information. This matches anchor 5 — every token earns its place and competence is assumed.

5 / 5

Actionability

Concrete shell commands are present ("aws sts get-caller-identity", "aws iam list-attached-user-policies", "aws s3 cp s3://bucket/terraform.tfstate -", "kubectl get pods -A -o json"), but many steps invoke undefined tools with no syntax or pointer — "iam_policy_audit", "tfstate_audit", "k8s_audit", "kg_add_node", "metadata_endpoints('aws')", "plan_attack_chains(promote=True)" — and "fetch JSON and iam_policy_audit" is vague with a "<me>" placeholder. Key execution details are missing, matching anchor 3 rather than 4's 'minor gaps'.

3 / 5

Workflow Clarity

Both workflows have a clear numbered sequence, but validation is implicit or absent: no check of the STS identity output, no verification of policy JSON before audit, and only an implicit checkpoint ("Confirmed creds") before graph updates. Sequence present with checkpoints missing or implicit matches anchor 3, not 4's 'most checkpoints present'.

3 / 5

Progressive Disclosure

The body is a concise overview with well-signaled one-level-deep references: a Playbooks table mapping each technique to its SKILL.md path, plus short inline workflows. No bundle files exist to inline, no nested references, and navigation is a single table — matching anchor 5's clear overview with well-signaled references.

5 / 5

Total

16

/

20

Passed

Description

71%Weight 40%Scale 1-5

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 specific, keyword-rich description that clearly states what the skill covers with concrete technique names. Its main weakness is that the 'when to use' guidance is relegated to metadata.when_to_use rather than the description itself, capping completeness and slightly limiting trigger discoverability.

Suggestions

Append an explicit trigger clause to the description, e.g. 'Use when the engagement touches cloud assets — AWS/GCP/Azure IAM, buckets, Kubernetes, or Terraform — or when an SSRF reaches a cloud metadata endpoint.'

Add missing natural synonyms (Azure, GCP, cloud security, SSRF, pod escape) so the description matches user phrasing beyond AWS-centric terms.

Distinguish the lane-catalog role (routing to per-technique playbooks) from the individual playbook skills to reduce overlap on shared trigger terms.

DimensionReasoningScore

Specificity

The description lists five concrete capabilities — "AWS IAM privesc, S3 takeover, k8s RBAC abuse, Terraform state leaks, cloud metadata pivoting" — matching the anchor for multiple specific concrete actions with comprehensive coverage. It is in noun-phrase (third person) form with no vague filler.

5 / 5

Completeness

The "what" is clear (a cloud exploitation lane covering five named techniques), but there is no "Use when..." clause in the description field — trigger guidance lives only in metadata.when_to_use. Per the guideline, a missing explicit 'when' clause caps completeness at 3; it has a clear 'what' with 'when' only weakly implied by "lane".

3 / 5

Trigger Term Quality

Strong natural keywords are present ("cloud", "AWS", "IAM", "S3", "k8s", "Terraform", "metadata", "privesc") that users would plausibly say, but common variations like "Azure", "GCP", "cloud security", or "SSRF" are absent. Good coverage with a few natural terms missing matches anchor 4, not the comprehensive synonym coverage of anchor 5.

4 / 5

Distinctiveness Conflict Risk

The niche is clear and the technique-level terms ("k8s RBAC abuse", "Terraform state leaks") are distinct, but as a lane-level catalog it shares trigger terms (IAM, S3, k8s) with the per-technique playbook skills it routes to, creating minor overlap risk with closely related skills. This fits anchor 4 better than 5's 'minimal conflict risk'.

4 / 5

Total

16

/

20

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.

Validation — 14 / 16 Passed

Validation for skill structure

CriteriaDescriptionResult

metadata_version

'metadata.version' is missing

Warning

metadata_field

'metadata' should map string keys to string values

Warning

Total

14

/

16

Passed

Repository

PurpleAILAB/Decepticon

Path

packages/decepticon/decepticon/skills/standard/cloud/SKILL.md

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.