palarran-security-egress

Topics in the source

• Egress and outbound data audit

Invocation: /palarran:security:egress {scope} [--plan]. Read the profile guide, dated sources, shared audit contract, global memory and decision memory before analysis. 1. Identify subject revision, target/environment, actor/tenant model, data classes and approved destinations/recipients. Read target instructions, .palarran/INDEX.md, decisions and locale. Reuse relevant local evidence; do not introduce remote memory or target access through the memory workflow. 2. Default to a local read-only source/configuration audit. With --plan, produce the evidence/verification plan without executing commands, resolving domains, calling services, querying targets or refreshing remote state. Live read-only configuration inspection requires a named target, permitted read role and concrete data scope. URL fetching, DNS/redirect/metadata probes, callbacks, scans, exploit attempts and production exports are separate test/execution scope even when described as harmless. 3. Draw two journeys: input → URL/client/resolver/proxy → destination, and data source → authorization/projection → response/export/integration recipient. Include jobs/retries, cache/streaming paths and browser third parties. A netwo

This is an excerpt. Refer to the complete source and its current requirements.