Back to search
Top Contributors
39
commits
Synthclaw coagent is lightweight and python based alternative to openclaw. On openclaw, you setup everything. On Synthclaw-coagent, the agent setups everything for you.
Stars
1
Forks
0
Watchers
1
Open Issues
0
Overall repository health assessment
No package.json found
This might not be a Node.js project
39
commits
Sanitize final replies and expose raw JSON detection in /task
1b2dd78View on GitHubfix: make /stop and /task responsive during running requests
8dff3c6View on GitHubrefactor: split Anthropic/OpenAI providers from Gradient and add Gemini/Grok Gradient candidates
17d080cView on GitHubchore: add /task live status telemetry and strict long-term memory threshold
1fbef68View on GitHubfeat: context cache + auto durable long-term memory extraction
788e876View on GitHubfeat: stall-safe checkpoint resume + interactive provider key setup with validation
8b8a745View on GitHubfeat: add OpenRouter/GitHub providers, provider-key chat command, provider-first /models UI
cb19015View on GitHubtune: larger checkpoints, stronger 429 handling, less aggressive anti-loop
41c6f5bView on GitHubfix: suppress tool markup in anti-loop and checkpoint after repeated duplicate calls
36f1812View on GitHubfix: parse tokenized tool-call blocks and prevent raw tool payload leakage
33a78b5View on GitHubfeat: task checkpointing Continue button for long tasks, limited LLM retry
8153e97View on GitHubfeat: split /usage from /models, infinite LLM retry, fix model button edit
e31eb5fView on GitHubfeat: inline button model switching + educated pricing estimates
adc8824View on GitHubfeat: per-model token usage tracking + pricing in /models (verified DO prices, ? for unconfirmed)
be1b3f4View on GitHubfeat: expand AVAILABLE_MODELS to all 32 DO inference models (Anthropic, OpenAI, DO-hosted); group /models output by provider
4639638View on GitHub