Collapse the per-type observed-media dispatch into one platform-agnostic
cache_media_bytes() helper in gateway/platforms/base.py. Any adapter can now
hand it raw attachment bytes + a filename/MIME hint; it classifies against the
shared MIME registries, routes to the right cache_*_from_bytes helper,
sandbox-translates the path, and returns a CachedMedia with a ready
context_note(). Telegram's observed-group path shrinks to: size-gate, download,
call the helper, annotate. Also dedupes the addressed-media type ladder into
_media_message_type().
Net: contributor's Telegram-only +595 LOC becomes a +210/-32 production change,
with the reusable primitive available to Discord/Slack/Signal/etc.
Co-authored-by: Glucksberg <markuscontasul@gmail.com>
- Add .zip to SUPPORTED_DOCUMENT_TYPES so gateway platforms (Telegram,
Slack, Discord) cache uploaded zip files instead of rejecting them.
- Add get_cache_directory_mounts() and iter_cache_files() to
credential_files.py for host-side cache directory passthrough
(documents, images, audio, screenshots).
- Docker: bind-mount cache dirs read-only alongside credentials/skills.
Changes are live (bind mount semantics).
- Modal: mount cache files at sandbox creation + resync before each
command via _sync_files() with mtime+size change detection.
- Handles backward-compat with legacy dir names (document_cache,
image_cache, audio_cache, browser_screenshots) via get_hermes_dir().
- Container paths always use the new cache/<subdir> layout regardless
of host layout.
This replaces the need for a dedicated extract_archive tool (PR #4819)
— the agent can now use standard terminal commands (unzip, tar) on
uploaded files inside remote containers.
Closes: related to PR #4819 by kshitijk4poor
- Sanitize filenames in cache_document_from_bytes to prevent path traversal (strip directory components, null bytes, resolve check)
- Reject documents with None file_size instead of silently allowing download
- Cap text file injection at 100 KB to prevent oversized prompt payloads
- Sanitize display_name in run.py context notes to block prompt injection via filenames
- Add 35 unit tests covering document cache utilities and Telegram document handling
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
