feat: add image sending and receiving support to all claw channels#58
Open
varshinivij wants to merge 8 commits intoaristoteleo:mainfrom
Open
feat: add image sending and receiving support to all claw channels#58varshinivij wants to merge 8 commits intoaristoteleo:mainfrom
varshinivij wants to merge 8 commits intoaristoteleo:mainfrom
Conversation
Collaborator
|
The scope of these changes looks well-contained. Have you tested this across all channels? If the testing is fully covered, we are good to merge |
Collaborator
|
let me know if you think its good to merge. |
zqbake
reviewed
Apr 3, 2026
|
|
||
| # ── Auto-start configured Claw channels ───────────────────────────── | ||
| try: | ||
| from pantheon.claw import ClawConfigStore |
Collaborator
There was a problem hiding this comment.
pantheon claw alreay worked without this?
Collaborator
|
@varshinivij please fixed all test so that we can merge ur PR |
Wire end-to-end image support through the claw messaging gateway. Inbound images from all 7 channels (Telegram, Discord, Slack, WeChat, Feishu, QQ, iMessage) are now downloaded, encoded as base64 data-URIs, and sent to the chatroom as multimodal _llm_content. Outbound images from tool results (e.g. Python plots) are collected via a new image-aware step callback and delivered back to users through each platform's native image API. Includes 14 new tests covering the full round-trip pipeline. Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
… send generated images image_gen.py returned file paths but channel callbacks expected base64_uri in raw_content. Added base64_uri + hidden_to_model to all three generation methods, matching the pattern used by python_interpreter. Also fixed runtime to handle base64_uri as list, hoisted notebook image URIs, and added claw auto-start support. Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
…preter to leader - Fix _snapshot_images() to use _get_effective_workdir() so images created in workspace directories are detected and sent through claw channels - Add python_interpreter toolset to leader agent in default team template Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
…ut directory Instead of relying on each toolset to set base64_uri individually, add a chatroom-level pre/post snapshot that catches images saved to a designated .pantheon/images/ directory. Agents are instructed via system prompt to save all generated images there, keeping the scan cheap (single flat directory). - Add shared image_detection utility (snapshot, diff, encode) - Chatroom.chat() creates .pantheon/images/, snapshots before/after execution, and emits synthetic step messages for newly detected images - Shell toolset: remove Python-only guard so all commands trigger detection, refactored to use shared utility - Claw runtime: deduplicate images in make_image_step_callback - Agent: inject <image_output_constraint> when image_output_dir is set Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
The chatroom-level .pantheon/images/ catch-all handles universal image detection. Shell-level snapshot should stay targeted to Python commands only to avoid unnecessary I/O on every shell invocation. Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Downloaded images from Slack were passed as raw bytes to OpenAI without format normalization, causing "unsupported image" errors. Now all Slack images are re-encoded through PIL to guaranteed PNG/JPEG output. Also improved error logging for image upload failures. Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Add shared md_to_slack, md_to_telegram (MarkdownV2), and md_to_plain converters in runtime.py. Apply per-channel: Slack gets mrkdwn, Telegram gets MarkdownV2 with proper escaping and fallback, plain-text channels (Feishu, WeChat, QQ, iMessage) get stripped markdown. Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
30 new tests covering md_to_slack, md_to_telegram (MarkdownV2), and md_to_plain: bold/italic/strikethrough conversion, link handling, header rendering, list formatting, code block preservation, special char escaping, edge cases (empty strings, plain text), and full message round-trips. Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
3 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Wire end-to-end image support through the claw messaging gateway. Inbound images from all 7 channels (Telegram, Discord, Slack, WeChat, Feishu, QQ, iMessage) are now downloaded, encoded as base64 data-URIs, and sent to the chatroom as multimodal _llm_content. Outbound images from tool results (e.g. Python plots) are collected via a new image-aware step callback and delivered back to users through each platform's native image API. Includes 14 new tests covering the full round-trip pipeline.