hygiene

Use when enforcing defensive coding practices that treat all data as untrusted — regardless of source. Covers sanitization, canonicalization, encoding, and validation of inputs AND outputs at every component boundary, including internal databases, caches, message queues, and inter-service calls. USE FOR: data hygiene, sanitization, canonicalization, normalization, output encoding, trust boundaries, defense in depth, zero trust data handling, internal data validation, component boundary security DO NOT USE FOR: specific injection attack patterns (use input-validation or owasp), cryptographic controls (use cryptography), API-level security headers (use api-security)

インストールへ移動

リポジトリ: Tyler-R-Kendrick/agent-skills
ソースの最終更新活動: 2026年2月11日 05:14
検出された SKILL.md の言語: 英語
スター: 10
フォーク: 4

インストール方法

デフォルトでは、最初にソースを確認する Prompt が選択されています。直接コマンドに切り替えるか、ローカルコピーをダウンロードすることもできます。

Codex または Claude でインストールこの Prompt をコピーして Codex、Claude、または他のアシスタントに貼り付けると、Skill ページを確認してインストールできます。

直接コマンドでは確認用 Prompt が省略されます。実行前にソースを確認してください。

npx skills add https://github.com/Tyler-R-Kendrick/agent-skills --skill hygiene

コマンドは1行のまま表示されます。コピー前に横へスクロールして全体を確認してください。

ローカルで確認しますか？SkillsMP が現在取得できるファイルをダウンロードできます。

ソースファイルを確認

インストールを決める前に、SKILL.md と SkillsMP に表示されている付属ファイルをお読みください。

ファイルエクスプローラー

16 ファイル

SKILL.md

ソースの指示 · 読み取り専用プレビュー

検索に戻るインストールに戻る

improve

Use when producing agent/LLM evals, synthetic simulation data, or self-improvement pipelines for prompts, code, skills, agents, harnesses, and workflows. Covers AgentEvals/AgentV, Agent Skills evals, ASSERT, GEPA, Trace, VISTA, Agent Lightning, SkillOpt, Simula-style data design, progressive disclosure, deterministic workspaces, and release evidence. USE FOR: eval creation, EVAL.yaml, AgentEvals, AgentV, evals.json, ASSERT, judge-traces, behavior taxonomy, judges, graders, rubrics, synthetic data, simulation data, Simula, QDC, source-grounded generation, prompt optimization, agent improvement, skill improvement, harness hardening, progressive disclosure, deterministic workflows, GEPA, Trace, VISTA, Agent Lightning, SkillOpt DO NOT USE FOR: ordinary unit/integration tests without AI quality criteria (use testing), refactoring without eval or trace feedback (use refactor), generic Agent Skills packaging without eval or improvement work (use agent-skills)

2026-06-09

Use when working with AI agent protocols, standards, interoperability specifications, evaluation contracts, synthetic simulation data, improvement pipelines, and agent steering workflows. Covers MCP, A2A, ACP, Agent Skills, AGENTS.md, ADL, Improve, x402, AP2, MCP Apps, cagent, and learn. USE FOR: agent protocol selection, comparing MCP vs A2A vs ACP, understanding agent standards ecosystem, choosing payment protocols, choosing eval standards, choosing improvement techniques, choosing synthetic data simulation techniques, steering from user feedback DO NOT USE FOR: specific protocol, eval, or improvement implementation details (use the sub-skills: mcp, a2a, acp, improve, learn, x402, etc.)

2026-06-09

learn

Use when a user corrects, rejects, edits, or redirects an LLM/agent response and the correction should become a reusable reasoning strategy. Converts feedback into generalized learnings for ~/.agents/STEERING.md with linked RDF/Turtle evidence. USE FOR: user corrections, preference feedback, rejected agent behavior, reasoning strategy updates, steering file maintenance DO NOT USE FOR: storing task facts (use memory), ordinary skill authoring (use agent-skills), project instruction files unrelated to feedback (use agents-md)

2026-06-08

このリポジトリのすべてを見る

name	hygiene
description	Use when enforcing defensive coding practices that treat all data as untrusted — regardless of source. Covers sanitization, canonicalization, encoding, and validation of inputs AND outputs at every component boundary, including internal databases, caches, message queues, and inter-service calls. USE FOR: data hygiene, sanitization, canonicalization, normalization, output encoding, trust boundaries, defense in depth, zero trust data handling, internal data validation, component boundary security DO NOT USE FOR: specific injection attack patterns (use input-validation or owasp), cryptographic controls (use cryptography), API-level security headers (use api-security)
license	MIT
metadata	{"displayName":"Hygiene","author":"Tyler-R-Kendrick"}
compatibility	claude, copilot, cursor
references	[{"title":"OWASP Input Validation Cheat Sheet","url":"https://cheatsheetseries.owasp.org/cheatsheets/Input_Validation_Cheat_Sheet.html"},{"title":"OWASP Cheat Sheet Series","url":"https://cheatsheetseries.owasp.org"}]

Source	Why It Can't Be Trusted	Example Threat
Database	Rows may have been inserted by a compromised service, SQL injection, or legacy code without validation	Stored XSS: HTML in a database field rendered without encoding
Cache (Redis, Memcached)	Cache poisoning, stale data, no authentication by default	Attacker writes malicious payload to shared cache key
Message Queue	Any producer can publish; messages may be replayed or tampered	Deserialization attack via crafted message payload
Internal API / Microservice	Service may be compromised, misconfigured, or returning unexpected data	Downstream service returns user-controlled data without sanitization
Configuration Store	Config values may be modified by operators or injected via environment	Command injection through a config value used in a shell call
File System	Files may be written by other processes, users, or symlink attacks	Path traversal via a filename read from a temp directory
Logs (when read back)	Log entries may contain injected content	Log injection leading to SIEM manipulation or log4shell-style attacks

Strategy	What It Does	When to Use
Stripping	Removes disallowed characters or tags entirely	HTML tags from plain-text fields, control characters from strings
Escaping	Replaces special characters with safe equivalents	HTML entities, SQL parameterization, shell escaping
Allowlisting	Accepts only values matching a known-good pattern	Enum fields, expected formats (UUID, ISO date, email)
Type coercion	Converts data to the expected type	Parse string to integer, parse to date object

Technique	Example	What Canonicalization Does
Unicode normalization	`\uFF41` (fullwidth 'a') vs `a`	Normalize to NFC/NFD before comparing
URL encoding	`%2e%2e%2f` = `../`	Decode URL encoding before path validation
Double encoding	`%252e` → `%2e` → `.`	Decode iteratively until stable
Case variation	`SELECT` vs `SeLeCt`	Lowercase before matching keywords
Path normalization	`/app/../etc/passwd`	Resolve to canonical absolute path (`/etc/passwd`)
Null bytes	`file.php%00.jpg`	Strip null bytes before extension check
Homoglyph substitution	`pаypal.com` (Cyrillic 'а')	Normalize to ASCII or punycode
Whitespace injection	`admin\t` vs `admin`	Trim and normalize whitespace

Output Context	Encoding Required	Why
HTML body	HTML entity encoding (`<` → `<`)	Prevents XSS
HTML attributes	Attribute encoding (quote special chars)	Prevents attribute injection
JavaScript	JS string encoding or `JSON.stringify()`	Prevents script injection
SQL	Parameterized queries (never string concat)	Prevents SQL injection
Shell commands	Avoid shell entirely; if unavoidable, use library escaping	Prevents command injection
URLs	URL-encode query parameters and path segments	Prevents parameter injection
XML	XML entity encoding	Prevents XXE and injection
JSON	Proper serialization (not string concatenation)	Prevents JSON injection
Log entries	Strip newlines and control characters	Prevents log injection
Email headers	Strip newlines from header values	Prevents header injection

#	Check	Applies To
1	Type validation — Is the data the expected type?	All inputs
2	Format validation — Does it match the expected pattern?	Strings, IDs, dates, emails
3	Range validation — Is it within acceptable bounds?	Numbers, dates, string lengths
4	Canonicalization — Is the data in its simplest canonical form?	Paths, URLs, Unicode strings
5	Allowlist check — Is the value in the set of permitted values?	Enums, status codes, roles
6	Sanitization — Have dangerous characters been removed or escaped?	All text inputs
7	Output encoding — Is data encoded for the destination context?	All outputs
8	Size limits — Is the data within acceptable size bounds?	Files, payloads, strings

hygiene

インストール方法

ソースファイルを確認

Hygiene — Trust No Data

Overview

Hygiene — Trust No Data

Overview

The Trust Boundary Model

Why Internal Sources Are Untrusted

Sanitization

Strategies

Sanitize Both Directions

Canonicalization

Common Bypass Techniques Prevented by Canonicalization

Canonicalization Order

Language Examples

Output Encoding

Context-Specific Encoding

Encoding Data From Internal Sources

Component Boundary Checklist

Anti-Patterns

Best Practices

Anti-Pattern	Why It's Dangerous	Correct Approach
"It's from our database, so it's safe"	Database can contain injected content from any entry point	Encode database values for the output context
"We validated it on the way in, so it's clean"	Output context may differ from input context; data may be modified in transit	Encode for the specific output context at point of use
"It's an internal API, so we trust it"	Internal services can be compromised or misconfigured	Validate response schema and sanitize values
"We use an ORM, so SQL injection is impossible"	Raw queries, dynamic column names, and ORDER BY bypass ORM protections	Always parameterize; review ORM escape hatches
"The cache just stores what we put in"	Cache can be poisoned via shared keys, race conditions, or unauthenticated access	Validate and encode cache values on read
"Config values are set by operators"	Environment variables and config stores can be compromised	Validate config values at startup; never pass raw to shell or template

hygiene

ソース情報

インストール方法

ソースファイルを確認

このリポジトリの他の Skills

このリポジトリの他の Skills

Hygiene — Trust No Data

Overview

Hygiene — Trust No Data

Overview

The Trust Boundary Model

Why Internal Sources Are Untrusted

Sanitization

Strategies

Sanitize Both Directions

Canonicalization

Common Bypass Techniques Prevented by Canonicalization

Canonicalization Order

Language Examples

Output Encoding

Context-Specific Encoding

Encoding Data From Internal Sources

Component Boundary Checklist

Anti-Patterns

Best Practices