技能 mantis-structural-index
📦
mantis-structural-index
內容修訂 r1 高風險 ⚙️ 外部命令📁 檔案系統存取
Build Snapshot-Aware Source Code Indexes
Large source trees make symbol navigation and caller discovery expensive. This skill guides structural indexing with reusable caches and coverage-aware queries while preserving lexical discovery.
支援: Claude Codex Code(CC)
安裝前請停止並請求確認。
請審查安裝計畫，並在變更檔案前取得使用者明確同意。
使用我的智慧代理安裝
將此請求複製給你的智慧代理，其中包含權威 Skill 頁面和清單。
智能體請求
Review the Skillstore skill "mantis-structural-index" from https://skillstore.io/skills/google-mantis-structural-index.md and its manifest at https://skillstore.io/api/skills/google-mantis-structural-index/manifest. Verify the artifact. Stop and obtain explicit user consent before installing or changing files.你的智慧代理仍應顯示其計畫，並依安全政策請求必要確認。
Agent 可讀資源
當 AI Agent、爬蟲或腳本需要乾淨脈絡、而不是讀取完整頁面時，請使用這些連結。
測試它
正在使用「mantis-structural-index」。 Illustrative request: index a live repository using only a heuristic backend.
預期結果:
- Snapshot provenance: unknown; a fresh build is required.
- Backend precision: heuristic; structural matches remain advisory.
- Returned artifacts: manifest, SQLite catalog, and query helper.
- Coverage summary distinguishes indexed, deferred, and failed files.
正在使用「mantis-structural-index」。 Illustrative request: find callers when the relevant partition is only partially indexed.
預期結果:
- Result status: no indexed callers; coverage is partial.
- This does not establish that the function has no callers.
- Run lexical discovery and retain all candidates alongside structural hints.
安全審計
高風險v1 • 2026/10/4 開啟版本化報告
All 225 static matches are documentation syntax or legitimate indexing operations, not the reported threats. Semantic review identifies forgeable helper-version checks and missing remote-query disclosure controls. No evidence found of malicious payloads, credential theft, or audit-directed prompt injection in the supplied skill.
1
已掃描檔案
747
分析行數
0
複核項目
0
已忽略的誤報
已確認的安全問題 (2)
高
Helper Reuse Relies on a Forgeable Version Marker
Existing executable helpers are reused after checking only the marker '# MANTIS_HELPER_VERSION = 5'. An attacker who can modify workspace helpers can retain that marker and substitute code executed during indexing or queries.
The instructions explicitly permit helper reuse based on a public version comment without requiring content integrity or trusted ownership. Exploitation requires workspace write access; no actual substituted helper was supplied.
中
Remote Queries Lack an Explicit Data-Disclosure Boundary
Remote query providers can be selected through MANTIS_STRUCTURAL_INDEX_URL or manifest configuration. Symbol and file queries may disclose repository metadata without required endpoint approval, transport protection, or a data-sharing policy.
The specification offers remote access to operations accepting symbol names and file paths but states no explicit endpoint or disclosure restrictions. No evidence found of a malicious destination or implemented transfer; this is a conditional design risk.
風險因素
⚙️ 外部命令 (50)
SKILL.md:14 SKILL.md:15 SKILL.md:21 SKILL.md:22 SKILL.md:30 SKILL.md:32 SKILL.md:34 SKILL.md:36 SKILL.md:41 SKILL.md:43 SKILL.md:45 SKILL.md:47 SKILL.md:49 SKILL.md:52 SKILL.md:54 SKILL.md:56 SKILL.md:58 SKILL.md:60 SKILL.md:62 SKILL.md:64 SKILL.md:66 SKILL.md:68 SKILL.md:70 SKILL.md:73 SKILL.md:74 SKILL.md:75 SKILL.md:78 SKILL.md:82 SKILL.md:83 SKILL.md:84 SKILL.md:93-132 SKILL.md:132-142 SKILL.md:142 SKILL.md:143-144 SKILL.md:144-147 SKILL.md:147-148 SKILL.md:148-149 SKILL.md:149 SKILL.md:150-153 SKILL.md:153 SKILL.md:154 SKILL.md:155 SKILL.md:157-158 SKILL.md:158-171 SKILL.md:171-172 SKILL.md:172 SKILL.md:174-175 SKILL.md:175-177 SKILL.md:177 SKILL.md:178-180
審計單位: codex
分享與引用此報告
分享版本化評估報告、中性徽章、嵌入卡片與引用。Skillstore 只報告證據，不判斷該 Skill 是否安全。
複製報告連結
https://skillstore.io/skills/google-mantis-structural-index/audits/1?utm_source=security_passport&utm_medium=share&utm_campaign=versioned_reportMarkdown 徽章
[](https://skillstore.io/skills/google-mantis-structural-index?utm_source=security_passport_badge)HTML 徽章
<a href="https://skillstore.io/skills/google-mantis-structural-index?utm_source=security_passport_badge"><img src="https://skillstore.io/badges/skills/google-mantis-structural-index/security.svg" alt="Skillstore security assessment" loading="lazy"></a>嵌入卡片
<iframe src="https://skillstore.io/embed/skills/google-mantis-structural-index.html" title="Skillstore Security Assessment" sandbox="allow-popups allow-popups-to-escape-sandbox" loading="lazy" referrerpolicy="no-referrer" width="420" height="180"></iframe>學術引用 (APA · BibTeX · CFF)
APA 引用
google. (2026). mantis-structural-index security audit report (audit version 1) [Author version unspecified]. Skillstore. https://skillstore.io/skills/google-mantis-structural-index/audits/1BibTeX 引用
@techreport{google-google-mantis-structural-index-2026,
author = {google},
title = {mantis-structural-index security audit report (audit version 1)},
institution = {Skillstore},
year = {2026},
number = {1},
url = {https://skillstore.io/skills/google-mantis-structural-index/audits/1},
note = {Author version unspecified}
}CITATION.cff
cff-version: 1.2.0
message: "If you use this Skill, cite its author and this versioned security audit report."
title: "mantis-structural-index security audit report (audit version 1)"
version: "unspecified"
type: report
authors:
- name: "google"
date-released: "2026-10-04"
url: "https://skillstore.io/skills/google-mantis-structural-index/audits/1"
identifiers:
- type: other
value: "skillstore:google-mantis-structural-index:audit:1"
description: "Skillstore immutable audit report identifier"
Skillstore 評分
評分依據 證據置信度: 中 55
架構
85
可維護性
87
內容
65
社群
83
規範符合度
你可以打造什麼
Navigate an Unfamiliar Codebase
Build source cross-references, resolve ambiguous symbols, and locate enclosing functions before reading implementation details.
Prioritize Security Research
Rank candidate call sites with structural hints while retaining every candidate found through repository-wide lexical discovery.
Reuse Indexing Work Across Snapshots
Reuse unchanged semantic units and report coverage, provenance, and deferred work for repeated analysis runs.
試試這些提示
Index the Current Repository
Build a local structural index for the current repository. Use unknown snapshot provenance, report coverage and backend precision, and preserve source files.
Resolve a Symbol and Its Callers
Resolve [symbol name] in [language or file]. Ask me to disambiguate matches, then list paginated caller hints with precision and coverage. Include lexical candidates.
Index a Pinned Snapshot
Index [snapshot root] using [snapshot ID] and [state root]. Verify provenance, reuse matching units, keep the snapshot read-only, and report deferred partitions.
Plan a Large Incremental Build
Index [snapshot root] locally with approved backends and verified helpers. Prioritize [target files] and changed units within [unit and byte limits]. Report reuse, remaining work, and coverage without excluding lexical candidates.
最佳實務
- Validate snapshot provenance and keep generated artifacts outside pinned source trees.
- Verify helper integrity before execution and approve any remote provider before sharing repository metadata.
- Inspect precision and coverage, disambiguate symbols, and combine structural results with lexical discovery.
避免
- Treating an empty structural query as proof that no callers exist.
- Reusing an unknown-snapshot index against a mutable live tree.
- Executing modified helpers solely because their version comment matches.
常見問題
Does this skill include executable indexing scripts?
No. The supplied file defines instructions and contracts. The agent generates builder and query helpers at runtime.
Can it run without a pinned snapshot?
Yes. It indexes the current directory with unknown provenance and rebuilds on every invocation.
Which indexing backends can it use?
It selects available backends per partition, including matched native indexes, compiler extractors, AST parsers, symbol tools, heuristics, and lexical fallback.
Does it replace repository-wide lexical searches?
No. Structural results are advisory. Consumers must retain lexical candidates and combine both result sets.
Can it analyze vulnerabilities or generate patches?
No. Its scope is source cross-reference data. Findings analysis, patching, reporting, and production build reachability are outside its scope.
Does it send repository information to remote services?
Local indexing does not require a remote provider. Optional remote queries can disclose symbol and file metadata; approve the destination before enabling them.