EASA Regulatory Search

Query EASA Easy Access Rules locally with exact reference lookup, full-text search, and semantic search. Use when answering EASA regulatory questions, retrie...

Security Scan

The skill's requests and instructions match its stated purpose (a local CLI for querying EASA Easy Access Rules); no unrelated credentials, unusual system access, or obfuscated code are present in the package.

Purpose & Capability

Name/description match the runtime instructions: the skill is an instruction-only wrapper around a local `claw-easa` CLI that performs exact-reference, full-text, and semantic search over EASA material. The files only require a local Python runtime and the upstream project; there are no environment variables, unrelated binaries, or unexpected credentials requested.

Instruction Scope

Runtime instructions tell the agent to call a local CLI (`claw-easa`) and to consult included reference docs. Commands include ingest/fetch/parse, search, and status checks. These actions are within the stated scope (downloading and indexing EASA ZIP/XML/FAQ pages and querying them). The instructions do allow crawling EASA FAQ pages and fetching archives (including an explicit `--url` override), which is consistent with ingestion behavior but worth noting as a network activity.

Install Mechanism

The skill package itself has no install spec (instruction-only), but the docs recommend cloning and running bootstrap scripts from https://github.com/dmoraine/clawEASA.git. Using a GitHub repo and provided scripts is expected for a Python CLI, but it does mean installation executes upstream scripts and installs Python packages (e.g., CPU Torch) from external sources. This is proportionate to a local runtime requirement but is the main place to review before running.

Credentials

The skill declares no required environment variables, credentials, or config paths. The operations described (web crawling, downloading EASA archives, parsing, local indexing) do not require secrets. No environment access appears unjustified by the skill purpose.

Persistence & Privilege

always:false and model invocation not disabled — normal for an agent-invocable skill. The package does not request persistent system-wide changes beyond advising copying the skill folder into the OpenClaw workspace; it does not modify other skills or request elevated privileges.

Scan Findings in Context

[no_code_to_scan] expected: The regex scanner had no code files to analyze because this is an instruction-only skill. That is expected; the runtime behavior depends on an external Python CLI repository referenced in the docs.

Assessment

This skill is coherent with its purpose, but it relies on an external Python project and runs network operations for ingestion. Before installing: (1) review the upstream GitHub repo and the bootstrap scripts (./scripts/bootstrap-local-runtime.sh and ./scripts/install-openclaw-skill.sh) to ensure you trust them, (2) run installs inside a dedicated virtualenv and not as root, (3) be cautious with the ingest command's `--url` option — only fetch archives from trusted sources, and (4) expect the skill to make outbound requests to EASA and to download/parse ZIP/XML FAQ pages for indexing.

Use the local claw-easa CLI from this repository.

Preferred commands:

• claw-easa lookup <REF> for exact references such as ORO.FTL.110
• claw-easa refs "<query>" for reference-oriented search
• claw-easa snippets "<query>" for cited text excerpts
• claw-easa hybrid "<query>" for mixed lexical + semantic retrieval
• claw-easa ask "<question>" for routed natural-language queries
• claw-easa status to verify corpus/index availability

Source-scoped search — use --slug <source> to restrict results to a specific source document. This is important when a broad query returns too many results from different sources, or when you know which source is most relevant:

• claw-easa refs "crew fatigue" --slug occurrence-reporting
• claw-easa snippets "crew fatigue" --slug occurrence-reporting
• claw-easa hybrid "fatigue reporting" --slug occurrence-reporting Use claw-easa sources-list to see available slugs.
• claw-easa sources-list to list ingested EARs and FAQ domains (supports --type ear|faq)
• claw-easa ear-discover to list Easy Access Rules available on the EASA website
• claw-easa ear-list to list built-in known source aliases
• claw-easa ingest fetch <slug> to download a source (use --url to bypass catalog)
• claw-easa ingest parse <slug> to parse a fetched source into the database
• claw-easa ingest diagnose <slug> to verify parser coverage against the source XML
• claw-easa ingest faq-all to ingest all EASA FAQs (crawls every sub-domain)
• claw-easa ingest faq <domain> to ingest FAQs for a specific domain
• claw-easa ingest faq-discover to list available FAQ domains on the EASA website

Answering rules:

• Prefer exact lookup when the user gives a regulation reference.
• Quote the retrieved text or excerpt before paraphrasing.
• Distinguish regulation text from AMC/GM/FAQ material.
• If retrieval is empty or ambiguous, say so explicitly instead of inferring.
• When the question targets a specific regulation domain (e.g. occurrence reporting, aircrew, air operations), use --slug to scope the search. This dramatically improves recall for items buried in long annexes.

Read these files only when needed:

• references/usage.md for repository-aware usage and installation notes
• references/runtime-setup.md when the local claw-easa CLI may not be installed yet
• references/easa-answering.md for answer format and evidence rules

Local install notes:

• The skill package lives under skill/claw-easa/ in the repository.
• For OpenClaw local installation, copy this folder into ~/.openclaw/workspace/skills/claw-easa/.
• Avoid symlinks that resolve outside the OpenClaw workspace.