CtrlK
BlogDocsLog inGet started
Tessl Logo
SkillPublic

sbom

Generate and manage Software Bill of Materials (SBOMs) for the OpenShell project. Covers SBOM generation with Syft, license resolution via public registries, and CSV export for compliance review. Trigger keywords - SBOM, sbom, bill of materials, license audit, license resolution, generate sbom, sbom csv, dependency license, supply chain, license scan.

Quality

81%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide
SecuritybySnyk

Low

Low-risk findings worth noting

SKILL.md
Quality
Evals
Security

Quality

Content

82%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.

The body is a strong, command-first reference: every workflow is backed by exact, executable commands with real paths, prerequisites are stated, and file pointers are one level deep and clearly tabulated. Weakest points are the fully redundant 'Quick Reference' section and the lack of an integrated verification checkpoint in the main pipeline (validation commands exist but are positioned as separate workflows).

Suggestions

Remove or drastically shrink the 'Quick Reference' section — it duplicates the exact commands already shown in Workflows 1–3.

Fold a verification checkpoint into Workflow 1 (e.g. a final step: 'Then run mise run sbom:check and confirm the unresolved-license report is acceptable before sharing the CSV').

Move the registry license-resolution table and image attestation prose into a short reference file (e.g. references/license-resolution.md) to keep SKILL.md as a lean overview.

DimensionReasoningScore

Conciseness

The body is efficient: dense tables (registries, output files, key files), copy-paste commands, and no concept explanations Claude already knows. The main trimmable redundancy is the 'Quick Reference' table, which repeats verbatim the commands already shown in Workflows 1–3 ('mise run sbom', 'mise run sbom:generate', etc.), plus slight overlap between Workflow 1's stage list and Workflow 2 — minor over-inclusion matching anchor 4. It is above anchor 3 because nearly every section delivers new, actionable information, and below anchor 5 because a whole section is redundant.

4 / 5

Actionability

Fully executable, copy-paste-ready commands throughout with real paths and flags: 'docker buildx imagetools inspect ghcr.io/nvidia/openshell/gateway:latest --format ...', 'tasks/scripts/verify-image-sbom.sh ... --require-cargo', 'mise x -- syft "file:deploy/docker/.build/..." -o cyclonedx-json', and 'uv run python deploy/sbom/resolve_licenses.py /path/to/external-sbom.json' — covering the common cases (full pipeline, per-stage, image inspection, external SBOMs), matching anchor 5.

5 / 5

Workflow Clarity

Sequences are clear: a one-command pipeline with its three named stages enumerated (Generate → Resolve → CSV), individual stages for debugging, an advisory CI check with its precondition stated ('Requires that SBOMs have already been generated'), and prerequisites up front. Validation exists ('verify-image-sbom.sh', 'sbom:check') but is presented as separate optional workflows rather than an explicit checkpoint in the main pipeline ('After running, the user can find...' lists outputs but doesn't tell Claude to verify them), which is the minor gap distinguishing anchor 4 from anchor 5.

4 / 5

Progressive Disclosure

Well-sectioned content with one-level-deep references to real project files clearly listed in the 'Key Files' table ('deploy/sbom/resolve_licenses.py', 'deploy/sbom/sbom_to_csv.py', 'tasks/sbom.toml'), and no nested references. It matches anchor 4 rather than 5 because detail like the registry resolution table and image-attestation pipeline prose is inlined in SKILL.md where a split into a reference file (or pruning) would make the overview leaner.

4 / 5

Total

17

/

20

Passed

Description

75%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.

The description is strong: it names the domain, three concrete capabilities with a specific tool (Syft), and provides an explicit trigger keyword list, so both 'what' and 'when' are answered. The main weakness is presentation — the 'when' guidance is a raw keyword list (with a redundant 'SBOM, sbom' case duplicate) instead of a scenario-oriented 'Use when...' clause, and it omits natural format terms like CycloneDX/SPDX.

Suggestions

Replace the 'Trigger keywords - ...' list with a proper 'Use when...' clause tied to user situations, e.g. 'Use when the user asks for an SBOM, bill of materials, or dependency/license audit, or mentions CycloneDX, SPDX, or exporting licenses to CSV.'

Include the natural ecosystem terms users would say — 'CycloneDX', 'SPDX', 'syft' — among the triggers, and drop the case-duplicate 'SBOM, sbom'.

Replace the generic 'Generate and manage' opener with the concrete actions (e.g. 'Generate CycloneDX SBOMs with Syft, resolve licenses via registry APIs, export CSV') and briefly mention image SBOM inspection/verification, which the skill body covers.

DimensionReasoningScore

Specificity

The description lists several specific actions with a concrete tool ('SBOM generation with Syft, license resolution via public registries, and CSV export for compliance review'), but 'Generate and manage' is generic and image SBOM inspection/verification from the body is not covered — minor gaps rather than comprehensive coverage, matching anchor 4. It does not reach anchor 5 because coverage of the skill's actual capabilities is incomplete, and it stays above anchor 3 because more than 1-2 concrete actions are explicitly named.

4 / 5

Completeness

Both 'what' and 'when' are present: the 'what' is concrete (generate, resolve, export) and the 'when' is explicit trigger guidance ('Trigger keywords - SBOM, sbom, bill of materials, license audit...'), so the missing-'Use when' cap at 3 does not apply. It sits at anchor 4 rather than 5 because the trigger guidance is a bare keyword list rather than scenario-framed 'Use when the user asks for...' phrasing, making the 'when' less explicit than anchor 5's concrete trigger phrases tied to user situations.

4 / 5

Trigger Term Quality

Good keyword coverage with natural phrases users would say ('sbom', 'bill of materials', 'license audit', 'generate sbom', 'supply chain'), but a few natural terms are missing — notably 'CycloneDX', 'SPDX', and 'syft' which users could plausibly say — matching anchor 4 rather than anchor 5's comprehensive synonym coverage. It is above anchor 3 because the list goes well beyond a couple of generic keywords and includes verb phrases and domain synonyms.

4 / 5

Distinctiveness Conflict Risk

The SBOM/compliance niche is distinct with specific triggers ('sbom', 'bill of materials', 'sbom csv'), but broader terms like 'supply chain', 'license audit', 'license scan', and 'dependency license' create minor overlap risk with security/vulnerability-scanning or license-compliance skills, matching anchor 4. It is above anchor 3 because the core triggers would not plausibly fire for unrelated skills, and below anchor 5 because several listed terms are not unique to this niche.

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

NVIDIA/OpenShell

Path

.agents/skills/sbom/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.