2 critical findings.

This skill exfiltrates environment secrets ($GH_TOKEN) plus 3 other issues listed below.

2 critical0 high2 medium8 rules passed

Why grade F?

score · 0 / 100

The current grade reflects 2 critical findings (any 2+ CRITs → F).

2 CRIT0 HIGH2 MED0 LOW

To reach a higher grade

D
Reach Dtarget score 30
Resolve 1 of your 2 CRIT findings — any single CRIT still keeps you at D.
C
Reach Ctarget score 55
Resolve all 2 CRIT findings.
B
Reach Btarget score 75
Resolve all 2 CRIT.
A
Reach Atarget score 95
Resolve all 2 CRIT.

Thresholds are documented at /docs/grading. Source-of-truth is the grade() function in @skillox/scanner.

Findings · ordered by severity

crit

Skill references secret env var $GH_TOKEN

The skill instructions reference `$GH_TOKEN`, which contains a credential. If an attacker can lure the agent into including this in an outbound URL or message, the credential leaks.

rule: env-var-harvestingline: 80CWE-200

▾

78```bash

79curl -s -X PUT \

80 -H "Authorization: Bearer $GH_TOKEN" \← references $GH_TOKEN — potential credential leak

81 -H "Accept: application/vnd.github+json" \

82 "https://api.github.com/repos/<owner/repo>/pulls/<pr_number>/merge" \

crit

Hardcoded secret detected: Anthropic / OpenAI API key (sk-…)

The skill body contains what looks like a literal `Anthropic / OpenAI API key (sk-…)`. Secrets must be read from the environment, never embedded in a published skill — embedded keys leak the moment the skill is mirrored or indexed.

rule: hardcoded-secretsline: 112CWE-798

▾

110

111```bash

112export OPENAI_API_KEY="sk-5Ds6eFbTEE1zu5fQ14F4FfB5892b419dB1BfC7292147B9Ef"← Anthropic / OpenAI API key (sk-…) — must be env-sourced, never inline

113export OPENAI_BASE_URL="http://152.53.52.170:3003/v1"

114CODEX_BIN="$HOME/.npm-global/bin/codex"

med

Arbitrary subprocess execution detected

The skill spawns subprocesses. Without a capability manifest declaring this, the skill could execute arbitrary commands.

rule: subprocess-executionline: 38CWE-78

▾

36### 步骤 3：并行启动 Claude Code

38用 exec(background=true) 并行启动多个 Agent，每个调用：← spawns a subprocess outside declared capabilities

40```bash

med

No capability manifest declared

The skill ships without a `manifest.yaml` or `capabilities` block in its frontmatter. Without a manifest, the runtime cannot enforce what this skill is permitted to do.

rule: no-manifest

▾

Scan another →Share

skillox.io/r/crawl-dyvbbg5ssexpsev9esg8w8x1