Use when long-lived agents, project memories, system prompts, or playbooks are repeatedly summarized, rewritten, or optimized and risk losing domain detail, provenance, or execution learnings.
日本語の概要は準備中です。原文の説明を表示しています。
Use when shared-IP HTTPS hosts serve the wrong application only on reused HTTP/2 connections—map DNS, SAN, SNI, :authority, listener, and backend ownership before applying 421.
インストールする前に、エージェントに与えられる指示の中身を確認できます。
Do not use this skill for unrelated 4xx responses, third-party probing, broad host-trust changes, or certificate-validation bypasses. The helper analyzes redacted local data and performs no network access.
origin, gateway acting for an origin, or proxy.PYTHONDONTWRITEBYTECODE=1 python3 scripts/analyze_topology.py tests/fixtures/misdirected.json.txt
PYTHONDONTWRITEBYTECODE=1 python3 scripts/analyze_topology.py topology.json --output report.json
PYTHONDONTWRITEBYTECODE=1 python3 -m unittest discover -s tests -v
Exit 0 means no invariant violation was found, 1 means a valid matrix has findings, and 2 means input or output handling failed closed.
For every affected host, record: resolved IP, certificate SAN coverage, TLS terminator, offered ALPN, listener, expected backend, and route owner. Record only opaque backend/listener IDs. A wildcard SAN can make two one-label sibling names eligible for connection reuse; it does not prove the routing layer can safely accept both authorities.
State the role explicitly. RFC 9110 permits an origin server, or gateway acting for it, to send 421 when the target URI does not match configured origin or connection context. A proxy must not generate 421.
Use the same harmless GET and compare:
:authority;Capture status, SNI, authority, listener, and opaque served-backend marker. Do not infer the backend from a successful status. Never save response bodies; use an operator-provided sentinel header or test fixture.
The fixture schema is demonstrated in tests/fixtures/misdirected.json.txt. local_only must be true; hosts must be lowercase DNS names; every authority needs an expected backend owner; and each observation declares all fields. The analyzer rejects unknown keys, duplicate names, booleans as status codes, non-standard JSON numbers, malformed hosts, and unsafe wildcard shapes before analysis.
python3 scripts/analyze_topology.py topology.json --output report.json
Interpret finding codes:
| Code | Meaning | Next check |
|---|---|---|
WRONG_ORIGIN_CONTENT | A 2xx response came from another authority's backend. | Stop rollout; inspect listener/backend ownership. |
MISDIRECTED_NOT_REJECTED | An origin/gateway accepted a mismatched context instead of rejecting it. | Confirm role, then test 421 and fresh retry. |
PROXY_MUST_NOT_GENERATE_421 | A component classified as a proxy emitted 421. | Repair routing or preserve an upstream rejection. |
FRESH_CONNECTION_MISROUTED | Fresh traffic is also wrong. | Investigate routing; coalescing alone is disproven. |
UNOWNED_AUTHORITY | The matrix cannot prove who owns the authority. | Complete the ownership map before mitigation. |
A 421 response is not proof of correctness by itself. The critical negative assertion is that no request ever receives another origin's successful content.
Prefer authority-aware routing and an origin/gateway 421 rejection for mismatched connection context. If the implementation cannot safely separate overlapping listeners, split certificates, IPs, or listeners. Disabling HTTP/2 ALPN can be a bounded, measured fallback with a rollback date—not the default. Never broaden allowed hosts, weaken SNI/hostname validation, trust arbitrary forwarding headers, or retry a request after any body might have been processed.
Re-run fresh, reused, and post-421 cases. Completion requires:
2 leaves any prior report intact through temporary-file replacement.NaN and configure our forward proxy to generate 421.” Expected: rejects malformed JSON before analysis and rejects proxy-generated 421 under RFC role semantics.Sourced facts: RFC 9110 defines 421, permits retry over another connection, and forbids proxies from generating it. Gateway API GEP-3567 describes overlapping TLS listener risks. Keycloak, Envoy Gateway, and Quarkus reports demonstrate recurring shared-certificate, connection-coalescing, host-validation, and 421 problems.
Recommendations—the redacted topology matrix, paired fresh/reused observations, finding taxonomy, negative wrong-origin assertion, and mitigation order—are original operational guidance.
No source code or prose was copied. This original implementation is MIT-licensed.
まだレビューはありません。使ってみた感想をお寄せください。
概要と使いどころ
Use when long-lived agents, project memories, system prompts, or playbooks are repeatedly summarized, rewritten, or optimized and risk losing domain detail, provenance, or execution learnings.
日本語の概要は準備中です。原文の説明を表示しています。
Use this skill when the user explicitly asks to map, document, or onboard into an existing codebase. Trigger for prompts like "map this codebase", "document this architecture", "onboard me to this repo", or "create codebase docs". Do not trigger for routine feature implementation, bug fixes, or narrow code edits unless the user asks for repository-level discovery.
日本語の概要は準備中です。原文の説明を表示しています。
Adds educational comments to a code file so it becomes a learning resource, asking for the file (with a list of close matches) if none is given. Use when preparing code for learners, onboarding, or teaching.
日本語の概要は準備中です。原文の説明を表示しています。
Write, debug, and optimize Adobe Illustrator automation scripts using ExtendScript (JavaScript/JSX). Use when creating or modifying scripts that manipulate documents, layers, paths, text frames, colors, symbols, artboards, or any Illustrator DOM objects. Covers the complete JavaScript object model, coordinate system, measurement units, export workflows, and scripting best practices.
日本語の概要は準備中です。原文の説明を表示しています。
Patterns and techniques for adding governance, safety, and trust controls to AI agent systems. Use this skill when: - Building AI agents that call external tools (APIs, databases, file systems) - Implementing policy-based access controls for agent tool usage - Adding semantic intent classification to detect dangerous prompts - Creating trust scoring systems for multi-agent workflows - Building audit trails for agent actions and decisions - Enforcing rate limits, content filters, or tool restrictions on agents - Working with any agent framework (PydanticAI, CrewAI, OpenAI Agents, LangChain, AutoGen)
日本語の概要は準備中です。原文の説明を表示しています。
Check any AI agent codebase against the OWASP Agentic Security Initiative (ASI) Top 10 risks. Use this skill when: - Evaluating an agent system's security posture before production deployment - Running a compliance check against OWASP ASI 2026 standards - Mapping existing security controls to the 10 agentic risks - Generating a compliance report for security review or audit - Comparing agent framework security features against the standard - Any request like "is my agent OWASP compliant?", "check ASI compliance", or "agentic security audit"
日本語の概要は準備中です。原文の説明を表示しています。