菜单
Review a code diff or corpus for security vulnerabilities grounded in the supplied threat model. Reads code semantically, forms hypotheses, verifies them, and emits scored findings as a single JSON code block. Deny-by-default: only the allowed_tools below are permitted.
用 Codex 或 Claude 帮你安装 复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。
基于 SOC 职业分类
Big-Sleep variant analysis. Given a recently-fixed bug shape (the lines added/removed in a patch), find more code in the repository that has the same shape and is likely to carry the same defect. Reports variants linked to their seed. Deny-by-default: only the allowed_tools below are permitted.
Given an architecture description or diagram, produce a STRIDE threat model plus attack trees as YAML. Use when the user asks to threat model a system.
Triage a candidate finding by deciding whether the vulnerable code is reachable from an untrusted entry point, producing the reachability axis (0-1) that feeds asec-confidence. Use when the user asks to triage a finding or assess exploitability.
Review a git diff for vulnerabilities by running Semgrep and CodeQL, then emit structured findings. Use when the user asks for a security review of a diff/PR.
Given an architecture description, diagram, or repo, produce a Phase-Zero STRIDE threat model plus attack trees as YAML conforming to asec-threat-model schema. Use when the user asks to bootstrap or create a threat model for a system.
Critique an existing threat model for completeness and rigor — missing STRIDE categories, unjustified skips, weak mitigations, ungrounded likelihood/impact. Use when the user asks to review, critique, or red-team a threat-model.yaml.
| name | security-code-review |
| description | Review a code diff or corpus for security vulnerabilities grounded in the supplied threat model. Reads code semantically, forms hypotheses, verifies them, and emits scored findings as a single JSON code block. Deny-by-default: only the allowed_tools below are permitted. |
| allowed_tools | ["Read","Grep","Glob","Bash"] |
| argument_hint | [target-dir] |
Review the target corpus against the supplied threat model and emit machine-parseable
findings. Read code semantically — do not rely on a single regex. Prefer the
least-privilege tool: Grep/Glob to locate sinks, Read to confirm data flow, and
Bash only for read-only inspection.
Use when reviewing a pull request, diff, or small corpus against a known threat model to surface and verify security findings (CWE-89, CWE-79, CWE-22 are the focus here).
%, .format(), or + concatenation.cursor.execute(...) / session.execute(...) sink without
a parameterized placeholder (? / %s bind, or an ORM expression).error when the sink is reachable from an HTTP handler.request.args, request.form, or
similar directly into markup (<div>...</div>), bypassing an autoescaping template.error for reflected, browser-rendered output.os.path.join,
pathlib, or string concatenation with no normalization/containment check.open/read/served. A missing os.path.realpath
containment check under the base directory is the tell.error when the traversal feeds a file read/serve sink.Grep/Glob; confirm the data flow with Read.--network=none, E3).Emit findings as a single JSON code block containing an array of objects, each with the keys below. Emit nothing else after the block:
[
{
"rule_id": "CWE-89",
"message": "SQL injection: request param interpolated into SELECT.",
"severity": "error",
"cwe": "CWE-89",
"uri": "src/api/users.py",
"start_line": 17,
"end_line": 18,
"snippet": "cursor.execute(f\"SELECT * FROM users WHERE name='{name}'\")"
}
]
severity is one of error | warning | note | none.uri is the target-relative path; start_line/end_line are 1-based.cwe mirrors rule_id (e.g. CWE-89).allowed_tools; the permission gate denies by default (E8).