技能 Hook Development
📦

Hook Development

v0.1.0 內容修訂 r1 高風險 ⚙️ 外部命令📁 檔案系統存取🔑 環境變數🌐 網路存取

建置 Claude Code Hooks

Hook 工作流程需要安全的驗證、清楚的事件，以及可重複的測試。此技能提供 Claude Code hook 設定的模式、腳本和遷移指引。

支援: Claude Codex Code(CC)
Maintain this Skill? Claim it ⚠️ 38 差

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 "Hook Development" from https://skillstore.io/skills/davila7-hook-development.md and its manifest at https://skillstore.io/api/skills/davila7-hook-development/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。

Markdown 詳情 GET /skills/davila7-hook-development.md 已簽名 manifest GET /api/skills/davila7-hook-development/manifest 已簽名 lockfile GET /api/skills/davila7-hook-development/lockfile

測試它

正在使用「Hook Development」。 建立一個會封鎖高風險檔案寫入的 hook。

預期結果:

  • 針對 Write 和 Edit 操作的 PreToolUse hook 計畫。
  • 針對系統路徑、遍歷和認證檔案的安全準則。
  • 常見案例中預期的 allow、ask 和 deny 行為。

正在使用「Hook Development」。 將我的 Bash validation hook 遷移為 prompt hook。

預期結果:

  • 保留原始 policy goal 的 prompt hook 設計。
  • 已移除的 shell 邊界案例與新的模型相依風險之比較。
  • 啟用 hook 前用於測試的 rollout checklist。

正在使用「Hook Development」。 在發布外掛程式前測試 hook。

預期結果:

  • 針對 hook 設定、sample input 和 expected exit behavior 的驗證流程。
  • 需要測試的 timeout、invalid input 和 malformed output 案例清單。
  • 精簡的部署就緒摘要。

安全審計

高風險
v10 • 2026/7/19 開啟版本化報告

Of 308 static alerts, most are false positives caused by Markdown formatting, defensive checks, and documented shell examples. Confirmed risks include shell injection, unsafe temporary files, raw SQL construction, and unredacted audit logging. Recommended prompt and command validators also contain bypass and injection weaknesses.

11
已掃描檔案
2,781
分析行數
17
複核項目
0
已忽略的誤報

已確認的安全問題 (6)

高
SKILL.md:29references/migration.md:66
Prompt Decisions Consume Untrusted Instructions
Recommended prompt hooks interpolate tool input directly without delimiters or an explicit untrusted-data rule. Crafted commands can manipulate approval decisions.
Both recommended templates place attacker-controlled command text directly inside decision prompts. No injection-resistant framing or deterministic fallback is shown.
高
examples/validate-bash.sh:24-43
Destructive Command Denylist Is Bypassable
The Bash validator denies only exact substrings and approves every unmatched command. Equivalent destructive syntax or alternate tools bypass the hook.
The code returns success after several literal substring checks. Variants such as separated flags are not covered and reach the approval path.
高
examples/validate-write.sh:19-38
Protected Path Checks Ignore Symlinks
The write validator checks raw path strings without canonicalization, then approves other paths. Symlinked paths can reach protected or sensitive files.
The validator performs only substring and prefix comparisons. It never resolves the target or verifies that the canonical path remains within an allowed root.
高
references/advanced.md:294-295
Raw Hook Input Is Embedded in SQL
The database logging example places raw hook input inside a SQL statement. Crafted input can alter the query or expose database data.
The psql command inserts the complete input inside a single-quoted SQL literal. No parameter binding or SQL escaping is present.
高
references/advanced.md:353-358
Audit Log Stores Complete Hook Payloads
The audit example writes complete hook input to a persistent home-directory log. Tool arguments may contain source code, credentials, or user prompts.
The input is captured from stdin and appended without filtering or redaction. The documented log path is persistent and no restrictive permissions are set.
中
examples/validate-write.sh:19-34
Hook Output Allows JSON Injection
The write validator concatenates an unescaped tool path into JSON output. Quotes or control characters can corrupt or alter the decision response.
file_path comes from tool input and is inserted directly between JSON quotes. No JSON encoder escapes attacker-controlled characters.
能力複核項目 (17)

這些是真實的本機能力，對此技能而言可能是預期行為，因此需要複核，但不會按已確認的惡意行為計分。

高
references/advanced.md:358
Hidden file in home directory
echo "$timestamp | $USER | $tool_name | $input" >> ~/.claude/audit.log
The example persistently appends complete hook input to a hidden file in the user home directory. That payload can contain credentials, prompts, or source data.
高
references/advanced.md:358
Hidden file access
echo "$timestamp | $USER | $tool_name | $input" >> ~/.claude/audit.log
The example writes unredacted hook input into a hidden audit log. The documented access creates a real local disclosure and retention risk.
高
references/advanced.md:413
Temp directory access
export CLAUDE_PROJECT_DIR="/tmp/test-project"
The integration example uses a fixed shared temporary project path and later removes it recursively. Existing data at that path can be modified or deleted.
高
references/advanced.md:294
Database connection strings
psql "$DATABASE_URL" -c "INSERT INTO hook_logs (event, data) VALUES ('PreToolUse', '$input')" \
The example uses DATABASE_URL and inserts raw hook input into a database command. It exposes a connection secret to a fragile logging path with injection risk.
高
scripts/test-hook.sh:190
Shell command substitution
output=$(timeout "$TIMEOUT" bash -c "cat '$TEST_INPUT' | $HOOK_SCRIPT" 2>&1)
The runner interpolates user-supplied hook and input paths into bash -c. Shell metacharacters in either path can execute unintended commands.
中
references/advanced.md:97
Temp directory access
echo "$risk_level" > /tmp/hook-state-$$
The example writes state to a predictable PID-based path in a shared temporary directory. A precreated symlink can redirect or corrupt the write.
中
references/advanced.md:105
Temp directory access
risk_level=$(cat /tmp/hook-state-$$ 2>/dev/null || echo "unknown")
The example reads state from a predictable shared temporary path without ownership checks. Another local process can prepopulate the file and influence hook decisions.
中
references/advanced.md:177
Temp directory access
cache_file="/tmp/hook-cache-$cache_key"
The cache filename is deterministic and placed in a shared temporary directory without secure creation. Local file poisoning or symlink redirection is possible.
中
references/advanced.md:237
Temp directory access
echo "0" > /tmp/test-count-$$
The example creates a predictable PID-based state file in a shared temporary directory. It does not use exclusive creation or verify file ownership.
中
references/advanced.md:238
Temp directory access
echo "0" > /tmp/build-count-$$
The example creates another predictable PID-based state file in a shared temporary directory. A local attacker can redirect the write through a symlink.
中
references/advanced.md:250
Temp directory access
count=$(cat /tmp/test-count-$$ 2>/dev/null || echo "0")
The counter is read from a predictable shared temporary path without ownership or type checks. File poisoning can alter cross-event enforcement.
中
references/advanced.md:251
Temp directory access
echo $((count + 1)) > /tmp/test-count-$$
The counter is written to a predictable shared temporary path without secure creation. Symlink redirection can overwrite an unintended file.
中
references/advanced.md:259
Temp directory access
test_count=$(cat /tmp/test-count-$$ 2>/dev/null || echo "0")
The Stop hook trusts a predictable temporary counter file. Another process can spoof the value and bypass the documented test-enforcement decision.
中
references/advanced.md:323
Temp directory access
rate_file="/tmp/hook-rate-$$"
The rate-limit state uses a predictable shared temporary path. File poisoning or symlink replacement can bypass limits or redirect later writes.
中
references/advanced.md:419
Temp directory access
if [ -f "/tmp/session-initialized" ]; then
The test trusts a fixed marker path in a shared temporary directory. Another process can create that marker and produce a false successful result.
中
scripts/test-hook.sh:171
Temp directory access
export CLAUDE_PROJECT_DIR="${CLAUDE_PROJECT_DIR:-/tmp/test-project}"
The test runner supplies a fixed shared temporary project directory to the executed hook. Tests can unintentionally interact with preexisting files at that path.
中
scripts/test-hook.sh:173
Temp directory access
export CLAUDE_ENV_FILE="${CLAUDE_ENV_FILE:-/tmp/test-env-$$}"
The environment file uses a predictable PID-based path that the child hook writes and the runner later reads and removes. Symlink attacks are possible.

風險因素

⚙️ 外部命令 (50)
examples/load-context.sh:1 examples/validate-bash.sh:8 examples/validate-bash.sh:11 examples/validate-bash.sh:1 examples/validate-write.sh:8 examples/validate-write.sh:11 examples/validate-write.sh:1 references/advanced.md:36 references/advanced.md:37 references/advanced.md:63 references/advanced.md:81 references/advanced.md:92 references/advanced.md:93 references/advanced.md:96 references/advanced.md:105 references/advanced.md:125 references/advanced.md:174 references/advanced.md:175 references/advanced.md:176 references/advanced.md:181 references/advanced.md:244 references/advanced.md:245 references/advanced.md:248 references/advanced.md:250 references/advanced.md:251 references/advanced.md:259 references/advanced.md:273 references/advanced.md:274 references/advanced.md:291 references/advanced.md:304 references/advanced.md:305 references/advanced.md:319 references/advanced.md:320 references/advanced.md:324 references/advanced.md:327 references/advanced.md:328 references/advanced.md:335 references/advanced.md:353 references/advanced.md:354 references/advanced.md:355 references/advanced.md:367 references/advanced.md:368 references/advanced.md:388 references/advanced.md:396 references/advanced.md:414 references/advanced.md:462 references/advanced.md:470 references/advanced.md:35 references/advanced.md:55 references/advanced.md:74
📁 檔案系統存取 (45)
examples/validate-bash.sh:31 examples/validate-write.sh:26 references/advanced.md:358 references/advanced.md:358 references/advanced.md:105 references/advanced.md:181 references/advanced.md:250 references/advanced.md:259 references/advanced.md:281 references/advanced.md:295 references/advanced.md:97 references/advanced.md:105 references/advanced.md:177 references/advanced.md:237 references/advanced.md:238 references/advanced.md:250 references/advanced.md:251 references/advanced.md:259 references/advanced.md:323 references/advanced.md:413 references/advanced.md:419 references/migration.md:117 references/migration.md:165 references/patterns.md:268 references/patterns.md:306 scripts/README.md:86 scripts/README.md:89 scripts/test-hook.sh:155 scripts/test-hook.sh:227 scripts/test-hook.sh:34 scripts/test-hook.sh:35 scripts/test-hook.sh:40 scripts/test-hook.sh:50 scripts/test-hook.sh:51 scripts/test-hook.sh:63 scripts/test-hook.sh:64 scripts/test-hook.sh:75 scripts/test-hook.sh:76 scripts/test-hook.sh:87 scripts/test-hook.sh:88 scripts/test-hook.sh:171 scripts/test-hook.sh:173 scripts/validate-hook-schema.sh:32 SKILL.md:534 SKILL.md:550
🔑 環境變數 (1)
references/advanced.md:294
🌐 網路存取 (1)
SKILL.md:693
審計單位: codex 檢視審計紀錄 →
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。

開啟版本化報告
安全評估

複製報告連結

https://skillstore.io/skills/davila7-hook-development/audits/10?utm_source=security_passport&utm_medium=share&utm_campaign=versioned_report

Markdown 徽章

[![Skillstore security assessment](https://skillstore.io/badges/skills/davila7-hook-development/security.svg)](https://skillstore.io/skills/davila7-hook-development?utm_source=security_passport_badge)

HTML 徽章

<a href="https://skillstore.io/skills/davila7-hook-development?utm_source=security_passport_badge"><img src="https://skillstore.io/badges/skills/davila7-hook-development/security.svg" alt="Skillstore security assessment" loading="lazy"></a>

嵌入卡片

<iframe src="https://skillstore.io/embed/skills/davila7-hook-development.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

davila7. (2026). Hook Development security audit report (audit version 10) [Author version 0.1.0]. Skillstore. https://skillstore.io/skills/davila7-hook-development/audits/10

BibTeX 引用

@techreport{davila7-davila7-hook-development-2026, author = {davila7}, title = {Hook Development security audit report (audit version 10)}, institution = {Skillstore}, year = {2026}, number = {10}, url = {https://skillstore.io/skills/davila7-hook-development/audits/10}, note = {Author version 0.1.0} }

CITATION.cff

cff-version: 1.2.0 message: "If you use this Skill, cite its author and this versioned security audit report." title: "Hook Development security audit report (audit version 10)" version: "0.1.0" type: report authors: - name: "davila7" date-released: "2026-07-19" url: "https://skillstore.io/skills/davila7-hook-development/audits/10" identifiers: - type: other value: "skillstore:davila7-hook-development:audit:10" description: "Skillstore immutable audit report identifier"

比較變體

3 個可安裝變體

每位作者的技能都保留為獨立的可安裝項目。推薦變體依 Skillstore 證據排序。

為何此變體排名第一

Skillstore 使用量較高
anthropics 推薦

hook-development

Skillstore 評分 38
證據置信度 中
Skillstore 使用量 25
已更新

2026-09-09

anthropics

anthropics-hook-development

Skillstore 評分 38
證據置信度 中
Skillstore 使用量 23
已更新

2026-10-24

davila7 目前

davila7-hook-development

Skillstore 評分 38
證據置信度 中
Skillstore 使用量 17
已更新

2026-09-09

Skillstore 評分

評分依據 證據置信度: 中
68
架構
100
可維護性
87
內容
70
社群
65
規範符合度

你可以打造什麼

建立安全 Hooks

設計 PreToolUse hooks，在工具執行前審查檔案寫入、Bash 命令和敏感路徑。

遷移 Shell Hooks

將脆弱的 Bash 檢查轉換為能夠針對內容和邊界案例進行推理的 prompt hooks。

測試 Hook 行為

使用隨附的驗證與測試腳本，檢查 hook 設定、輸出、逾時和失敗行為。

試試這些提示

建立基本 Hook 
建立一個 Claude Code PreToolUse hook，用來驗證 Write 和 Edit 操作。說明 matcher、hook type 和預期的 decision output。
新增命令驗證 
為 Claude Code 設計一個 Bash command validation hook。封鎖破壞性命令，在特權命令前先詢問，並包含安全的 timeout settings。
遷移到 Prompt Hooks 
審查這段現有的 hook logic，並提出 prompt hook 版本。保留 security goal，並列出 migration risks。
建置 Hook 測試計畫 
為 Claude Code hook package 建立測試計畫。包含 schema validation、sample inputs、expected exits、timeout checks 和 failure cases。

最佳實務

  • 針對內容相關的安全決策，優先使用 prompt hooks；針對小型且確定性的檢查，使用 command hooks。
  • 引用 shell variables、驗證 JSON input，並為每個 command hook 設定嚴格的 timeouts。
  • 部署前，使用具代表性的 allow、ask、deny、timeout 和 malformed input 案例測試 hooks。

避免

  • 不要複製會在未經遮罩或使用者同意的情況下記錄原始 hook input 的範例。
  • 不要透過插入不受信任的 file paths 或 script names 來建構 shell commands。
  • 不要只依賴簡單的 deny lists 來偵測破壞性命令或敏感檔案。

常見問題

這項技能可以協助我建置什麼？
它可協助建置用於驗證、自動化、工作階段內容和完成檢查的 Claude Code hooks。
它著重於 prompt hooks 還是 command hooks？
兩者都涵蓋，但在支援時，建議使用 prompt hooks 來進行內容感知決策。
它可以驗證 hook 設定檔嗎？
可以。隨附的 schema validation 指引會檢查 events、matchers、hook types、commands、prompts 和 timeouts。
我可以直接在生產環境中使用這些範例嗎？
不可以。請審查、強化並測試這些範例，因為其中有幾個是教學模式，而不是可直接用於生產環境的控制措施。
Marketplace 列表提到了哪些工具？
此列表以 Claude、Codex 和 Claude Code 搜尋為目標，而技術內容則是關於 Claude Code hooks。
啟用 hooks 前應該測試什麼？
部署前，請測試 valid input、denied operations、malformed JSON、timeout behavior 和 output formatting。

開發者詳細資訊

作者

davila7

授權

MIT

作者版本

v0.1.0

Skillstore 修訂

r1

儲存庫

https://github.com/davila7/claude-code-templates/tree/main/cli-tool/components/skills/development/hook-development

引用

dbe0e719813583400773166a0621f8a9b8185c72

維護新鮮度

2026/7/25

使用量

9 次下載 · 211 次瀏覽

檔案結構

📁 examples/

📄 load-context.sh

📄 validate-bash.sh

📄 validate-write.sh

📁 references/

📄 advanced.md

📄 migration.md

📄 patterns.md

📁 scripts/

📄 hook-linter.sh

📄 README.md

📄 test-hook.sh

📄 validate-hook-schema.sh

📄 SKILL.md

davila7 的更多技能

查看全部

相關技能

查看全部