Item: agent-bom vulnerability intel
Rating: 7.3
Author: Implexa

agent-bom vulnerability intel

Use agent-bom to check package, SBOM, inventory, and agent dependency exposure against OSV, GitHub Security Advisories, NVD, EPSS, and CISA KEV with explicit...

copies: implexa run clawhub/agent-bom-vulnerability-intel

view source

installs

stars

karma

SkillRank score ↗

7.3/ 10

evaluated by implexa, claude-haiku-4-5 · 2026-05-26

agent-bom-vulnerability-intel routes vulnerability checks through a unified scanner, supporting package lookup, sbom scanning, and offline review against osv, github advisories, nvd, epss, and cisa kev. explicit data-boundary choices and guardrails preserve advisory provenance and protect sensitive identifiers.

structure

9.0

trigger phrases

8.0

procedure

7.0

edge cases

7.0

documentation

7.0

view original SKILL.md from clawhubclick to expand

---
name: agent-bom-vulnerability-intel
description: >-
  Use agent-bom to check package, SBOM, inventory, and agent dependency
  exposure against OSV, GitHub Security Advisories, NVD, EPSS, and CISA KEV
  with explicit data-boundary choices. Use when a user asks for CVE lookup,
  advisory intelligence, exploitability context, fix versions, GHSA/OSV/NVD
  enrichment, or package vulnerability triage.
version: 0.88.4
license: Apache-2.0
compatibility: >-
  Requires Python 3.11+ and agent-bom installed from this repository or PyPI.
  No credentials are required for basic public advisory lookups. Optional
  NVD_API_KEY and GITHUB_TOKEN values only raise provider rate limits.
metadata:
  author: msaad00
  homepage: https://github.com/msaad00/agent-bom
  source: https://github.com/msaad00/agent-bom
  pypi: https://pypi.org/project/agent-bom/
  openclaw:
    requires:
      bins:
        - agent-bom
      env: []
      credentials: none
    credential_policy: "Do not ask users to paste credentials. Optional NVD_API_KEY and GITHUB_TOKEN values may be present in the operator environment for rate limits, but their values must never be displayed, logged, or copied into prompts."
    optional_env:
      - NVD_API_KEY
      - GITHUB_TOKEN
    optional_bins: []
    emoji: "\U0001F6E1"
    homepage: https://github.com/msaad00/agent-bom
    source: https://github.com/msaad00/agent-bom
    license: Apache-2.0
    os:
      - darwin
      - linux
      - windows
    credential_handling: "No cloud or source-control credentials are needed. Advisory API tokens stay in the operator environment and are used only by agent-bom's existing advisory clients; do not echo or persist token values."
    data_flow: "Default package checks send package names, versions, ecosystems, PURLs, and CVE/advisory IDs to public advisory databases. Source code, raw config files, secrets, env values, and full scan reports are not sent to advisory providers. Use offline/cache-approved mode when private package names are sensitive."
    file_reads:
      - "operator-provided inventory JSON"
      - "operator-provided CycloneDX/SPDX SBOM files"
      - "local agent configuration paths only when the operator chooses a local scan"
    file_writes:
      - "operator-selected JSON/SARIF/report output path"
    network_endpoints:
      - url: "https://api.osv.dev/v1"
        purpose: "OSV package vulnerability lookup"
        auth: false
      - url: "https://api.github.com/advisories"
        purpose: "GitHub Security Advisories lookup; optional token only raises rate limits"
        auth: false
      - url: "https://services.nvd.nist.gov/rest/json/cves/2.0"
        purpose: "NVD CVSS, CWE, and publication metadata enrichment"
        auth: false
      - url: "https://api.first.org/data/v1/epss"
        purpose: "EPSS exploit probability enrichment"
        auth: false
      - url: "https://www.cisa.gov/sites/default/files/feeds/known_exploited_vulnerabilities.json"
        purpose: "CISA Known Exploited Vulnerabilities enrichment"
        auth: false
    telemetry: false
    persistence: false
    privilege_escalation: false
    always: false
    autonomous_invocation: restricted
---

# agent-bom-vulnerability-intel

Use this skill to answer vulnerability-intelligence questions through
agent-bom's existing scanners and canonical evidence model. Do not create
one-off OSV, GHSA, NVD, EPSS, or KEV clients in the agent session; route through
agent-bom so advisory provenance, aliases, severity gates, cache behavior,
redaction, and output schemas stay consistent.

## Modes

Start with the smallest mode that answers the user:

| Mode | Use When | Data Boundary |
|------|----------|---------------|
| `explain-only` | User wants to know what would be queried | No advisory calls |
| `check-package` | User names one package/version/ecosystem | Only that package identifier is queried |
| `scan-local` | User wants findings from local agents or a local inventory file | Local parse first; advisory calls use package identifiers only |
| `offline-review` | Private package names cannot leave the environment | Use local/cache-approved data only; disclose reduced coverage |
| `export` | User wants PR gate, SARIF, JSON, or audit evidence | Write only to an operator-selected path |

## Guardrails

- Ask before scanning a broad filesystem path or local agent configs.
- Do not paste or reveal `NVD_API_KEY`, `GITHUB_TOKEN`, package-registry
  credentials, cloud credentials, or env values.
- Do not send full source files, lockfiles, config contents, secrets, or scan
  reports to advisory providers. agent-bom extracts package identifiers first.
- Treat unknown or unresolvable versions as coverage gaps, not clean results.
- Preserve advisory provenance. Do not collapse OSV, GHSA, NVD, EPSS, and KEV
  into a single unlabelled severity.
- Do not modify dependencies or install fixes unless the user explicitly asks
  for a remediation workflow.

## Workflows

### Explain the Boundary

When the user asks "what leaves my environment?", answer before running:

```text
This lookup sends package identifiers (name, version, ecosystem/PURL) and CVE
IDs to public advisory databases. It does not send source code, raw configs,
secrets, env values, credentials, or full scan reports. Use offline-review if
private package names are sensitive.
```

### Check One Package

```bash
agent-bom check flask==2.0.0 --ecosystem pypi
```

Use this for quick triage and fix-version checks. If the package name belongs
to a private registry or internal project, use `explain-only` first and let the
operator decide whether the identifier may be queried externally.

### Scan a Canonical Inventory

```bash
agent-bom agents --inventory inventory.json --format json --output findings.json
```

Use this after an operator-pull adapter or discovery skill emits canonical
inventory. The inventory can stop at the file boundary; scanning is an explicit
operator handoff.

### Export for a PR Gate

```bash
agent-bom agents --inventory inventory.json --format sarif --output agent-bom.sarif
```

Use SARIF only when the user wants GitHub code-scanning or AppSec PR-gate
evidence. Keep JSON for local analysis and audit trails.

### Offline Review

If external advisory calls are not allowed, run with the project's offline or
cache-approved mode and say clearly that coverage depends on the locally
available vulnerability database. Do not call a clean offline result equivalent
to a fresh OSV/GHSA/NVD lookup.

## Output Rules

- Show CVE/GHSA/PYSEC aliases together when available.
- Include severity source, fix version, EPSS, KEV status, CWE, and advisory
  source chain when present.
- Separate "no vulnerabilities found" from "not enough data to evaluate."
- Keep raw credentials and credential-bearing URLs out of output, logs, prompts,
  SARIF locations, and exported reports.

don't have the plugin yet? install it then click "run inline in claude" again.

extracted implicit decision logic into explicit if-else branches, added edge cases (rate limits, timeouts, unresolvable versions, coverage gaps), documented all five external advisory apis as inputs with urls and purposes, expanded procedure steps with concrete command examples and error handling, and clarified output contract and outcome signals.

intent

use this skill to answer vulnerability-intelligence questions through agent-bom's existing scanners and canonical evidence model. route all cve lookups, advisory enrichment, exploitability context, fix-version queries, and package-triage requests through agent-bom rather than building one-off osv, ghsa, nvd, epss, or kev clients. this keeps advisory provenance, aliases, severity gates, cache behavior, redaction, and output schemas consistent across the session.

inputs

required

python 3.11 or higher installed on the operator's system
agent-bom binary installed from pypi or the clawhub repository (https://pypi.org/project/agent-bom/)

optional (increases rate limits and coverage)

NVD_API_KEY: nist nvd api token, stored in operator environment only, never logged or echoed
GITHUB_TOKEN: github api token, stored in operator environment only, never logged or echoed

data inputs, user-provided

package name and version (ecosystem: pypi, npm, maven, golang, nuget, cargo, etc.)
sbom file in cyclonedx or spdx json/xml format
operator-provided inventory json with canonical package identifiers
local agent configuration paths (ask before scanning)

external advisory apis (no auth required for public lookups)

osv database (https://api.osv.dev/v1): package vulnerability lookup
github security advisories (https://api.github.com/advisories): ghsa and ecosystem-specific advisories, optional token only raises rate limits
nist nvd (https://services.nvd.nist.gov/rest/json/cves/2.0): cvss, cwe, publication metadata enrichment
first epss (https://api.first.org/data/v1/epss): exploit probability enrichment
cisa kev (https://www.cisa.gov/sites/default/files/feeds/known_exploited_vulnerabilities.json): known actively exploited vulnerabilities enrichment

procedure

identify the user's question and select the minimal mode that answers it (see decision points below). do not default to broad scanning.
if the user asks about data boundaries first (e.g., "what leaves my environment?"), explain before executing: "this lookup sends package identifiers (name, version, ecosystem/purl) and cve ids to public advisory databases. it does not send source code, raw configs, secrets, env values, credentials, or full scan reports. use offline-review mode if private package names are sensitive."
for single-package lookup mode, execute:
```
agent-bom check <package-name>==<version> --ecosystem <ecosystem>
```
example: agent-bom check flask==2.0.0 --ecosystem pypi. output: cve/ghsa/pysec aliases, severity, fix versions, epss, kev status, and advisory source.
for sbom or inventory scan mode, accept the operator-provided file path and execute:
```
agent-bom agents --inventory <path-to-inventory.json> --format json --output <operator-selected-path>
```
the inventory must contain canonical package identifiers. do not ingest raw lockfiles or config files directly; extract identifiers first.
for pr-gate or export mode, use sarif output only if the user explicitly requests github code-scanning or appsec integration:
```
agent-bom agents --inventory <path> --format sarif --output agent-bom.sarif
```
keep json for local analysis and audit trails.
for offline or cache-approved mode, check operator environment for local vulnerability databases and run with cache constraints:
```
agent-bom check --offline-only <package-name>==<version> --ecosystem <ecosystem>
```
output: clearly state that coverage depends on locally available data and is not equivalent to a fresh osv/ghsa/nvd lookup.
enrich output with provenance: when results arrive, display cve/ghsa/pysec aliases together, include severity source (osv, ghsa, nvd), fix versions, epss scores, kev status, cwe identifiers, and advisory source chain.
handle empty or incomplete results: distinguish "no vulnerabilities found after advisory lookup" from "insufficient version data to evaluate." flag unresolvable versions as coverage gaps, not clean results.
write exported output only to operator-selected paths. do not persist credentials, credential-bearing urls, env values, or secrets in any exported file.

decision points

if the user asks "what data leaves my environment?"

output the data-boundary explanation (see procedure step 2) before making any advisory call.

if the user names a single package and version

use check-package mode (procedure step 3). do not scan the entire local filesystem.

if the user provides an sbom, inventory, or local agent scan

use scan-local mode (procedure step 4). extract package identifiers first; do not send raw source, lockfiles, or config contents to advisory providers.

if the user's package name suggests a private or internal registry (e.g., "my-org-internal-lib")

ask permission before querying public advisory databases. offer explain-only mode to show what would be sent.

if the user requests pr-gate, github code-scanning, or appsec integration

use export mode with sarif output (procedure step 5). do not export sarif unless explicitly requested.

if external advisory lookups are forbidden or the operator declares a private environment

use offline-review mode (procedure step 6). clearly state reduced coverage and do not claim equivalence to live osv/ghsa/nvd queries.

if nvd_api_key or github_token are present in the operator environment

use them to increase rate limits, but never log, echo, paste, or display their values in output, logs, or prompts.

if a package version cannot be resolved or is not found in any advisory database

report as a coverage gap, not a clean result. recommend broader ecosystem scanning or manual triage.

if rate limits are hit (429 http status)

back off and explain to the user that advisory providers have throttled the request. suggest retrying after a delay or providing api tokens to increase limits.

if network timeout or transient failure occurs

retry up to 2 times with exponential backoff (1 second, then 3 seconds). if the failure persists, report advisory api availability as degraded and offer offline mode.

output contract

success output is a structured advisory report containing

package identifier (name, version, ecosystem, purl)
cve/ghsa/pysec identifiers with aliases grouped together
severity (cvss v3.1 or v2.0 base score, source: osv, ghsa, or nvd)
epss percentile and exploit probability (from first epss api)
kev status (in cisa kev? yes/no, date added if yes)
cwe identifiers and description
advisory source chain (osv, ghsa, nvd, or local cache)
fix version or remediation guidance
json or sarif output written only to the operator-selected file path

output must not contain

env variable values, api tokens, or credentials
credential-bearing urls or api endpoints with auth parameters
source code, raw lockfiles, config contents, or scan reports
unlabelled or collapsed severity (preserve osv, ghsa, nvd, epss, and kev as distinct sources)

format options

--format json: structured advisory data for audit, local analysis, and integration. default.
--format sarif: github code-scanning and pr-gate integration only when requested.
--format text: human-readable summary (use only for cli output, not persistence).

outcome signal

the skill worked when

the user receives a complete advisory report with cve/ghsa identifiers, severity, fix versions, epss, and kev status for the package(s) they asked about.
the user can distinguish between "no vulnerabilities found" and "insufficient version data to evaluate" (coverage gaps are flagged).
exported reports (json or sarif) are written to the operator-selected path without credentials, secrets, or full scan artifacts.
if offline mode was used, the output clearly states that coverage depends on local cache and is not equivalent to a live external lookup.
if the user asked about data boundaries, they receive a clear explanation of what package identifiers are sent to advisory apis before any lookup occurs.
rate limits, network failures, and unresolvable versions are reported honestly, not hidden or reframed as clean results.

credits: original author msaad00 (https://github.com/msaad00/agent-bom). enriched for implexa quality standards.

agent-bom vulnerability intel

related skills

intent

inputs

procedure

decision points

output contract

outcome signal