{"version":3,"sources":["../../src/safeContent/index.ts"],"sourcesContent":["// Safe content rendering — non-executable MDX (TRUST_MODES_SPEC §5.1, AGENT_AUTHORING\n// §10). The host/SDK-owned renderer an INTERPRETER app uses to render untrusted\n// Markdown/MDX-syntax content as data, with **no evaluator anywhere in the pipeline**.\n// This is the mandatory terminal for the MDX-from-mount gate (§10 delta 1) — a shared\n// (multi-writer / M3) board or wiki entry MUST render through here, never compiled MDX.\n//\n// It is a **T2 tool, not a platform gate**: it stops content executing as code; it does\n// not certify anything to the host, and CO-4 data-fencing is still required for anything\n// that *reads* the content (the agent case).\n//\n// R3-279 (phase 3): the implementation now lives in `@immediately-run/safe-content`\n// (the S3 content-framework tier's first extraction). This module — and every sibling\n// subpath below — is a byte-compat RE-EXPORT: `api-snapshot.json` is unchanged, every\n// existing `@immediately-run/sdk/safeContent/*` import keeps working, and no consumer\n// changes anywhere.\nexport * from '@immediately-run/safe-content';\n"],"mappings":";;;;;;;;;;;;;;;AAAA;AAAA;AAeA,gCAAc,0CAfd;","names":[]}