ToolUniverse

mims-harvard/ToolUniverse

Branch: fix/biomodels-non-json-diagnostics · HEAD: 2daf3ec9 · origin/main: 3e981b10

clean working tree

Tracked PRs

PR	Title	Status	Checks
#169	docs: refresh contributing guide	MERGED UNKNOWN	Run Tests (3.12): pass
#170	fix: avoid unicode logging errors on non-UTF streams	MERGED UNKNOWN	Run Tests (3.12): pass
#171	fix: add CTD non-JSON response diagnostics	MERGED UNKNOWN	Run Tests (3.12): pass
#172	fix: add compact UniProt entry output	MERGED UNKNOWN	Run Tests (3.12): pass
#173	docs: expand Codex CLI MCP setup	MERGED UNKNOWN	Run Tests (3.12): pass
#176	fix: stabilize task manager listing	OPEN MERGEABLE	Run Tests (3.12): pass
#177	docs: add OpenCode manual MCP setup	OPEN MERGEABLE	Run Tests (3.12): pass
#178	docs: clarify MCPB client scope	OPEN MERGEABLE	Run Tests (3.12): pass
#179	fix: add DisGeNET empty-result diagnostics	OPEN MERGEABLE	Run Tests (3.12): pass
#180	fix: add BioModels non-JSON diagnostics	OPEN MERGEABLE	Run Tests (3.12): pass

Open Issues

• #175 DisGeNET tool returns empty results due to outdated API endpoint/auth
• #174 Biomodels related tools seems failed to work
• #155 Add a description to improve Dispatch discoverability
• #137 CTD tool failed
• #131 MCPB manifest invalid: server type 'uv' not supported by Claude Code plugin system
• #129 Feature Request: Support for AWS Bedrock Models
• #68 MCPB manifest uses invalid server type 'uv' - should be 'python'
• #63 OpenCode support?
• #58 Bug Report: OLS tool `ols_get_term_ancestors` returns empty terms array despite total_items > 0
• #56 Length of data from UniProt_get_entry_by_accession exceeds context window of many LLMs

Local Test

Tests not run. Refresh with --run-tests.

Recent Commits

2daf3ec9 (HEAD -> fix/biomodels-non-json-diagnostics, origin/fix/biomodels-non-json-diagnostics) fix: add BioModels non-JSON diagnostics
3e981b10 (origin/main, origin/HEAD) Merge pull request #173 from mims-harvard/docs/expand-codex-cli-setup
b9d7f77c Merge pull request #172 from mims-harvard/fix/uniprot-compact-entry
61b94feb Merge pull request #171 from mims-harvard/fix/ctd-non-json-diagnostics
a6a7bf67 Merge pull request #170 from mims-harvard/fix/windows-safe-logging

Qworld

mims-harvard/Qworld

Branch: docs/reproduction-defaults · HEAD: 64f444e · origin/main: 5700dd0

clean working tree

Tracked PRs

PR	Title	Status	Checks
#2	fix: derive package version from metadata	OPEN MERGEABLE	checks: none
#3	fix: avoid mutating generate inputs	OPEN MERGEABLE	checks: none
#4	fix: normalize CLI input shapes	OPEN MERGEABLE	checks: none
#5	ci: add test workflow	OPEN MERGEABLE	checks: none
#6	fix: normalize CLI and client input shapes	OPEN MERGEABLE	checks: none
#7	docs: add CLI batch and resume examples	OPEN MERGEABLE	checks: none
#8	docs: document reproduction defaults	OPEN MERGEABLE	checks: none

Open Issues

• #1 Could you provide the exact hyperparameters for HealthBench and HLE?

Local Test

Tests not run. Refresh with --run-tests.

Recent Commits

64f444e (HEAD -> docs/reproduction-defaults, fork/docs/reproduction-defaults) docs: document reproduction defaults
5700dd0 (origin/main, origin/HEAD, fork/main, fork/HEAD, main) Update README.md
7e972cb Embed video thumbnail in README
1e69918 Add video and demo badges to README
2cb2114 Add citation section to README

Medea

mims-harvard/Medea

Branch: docs/targetid-data-inputs · HEAD: 3f72428 · origin/main: d44d1d4

clean working tree

Tracked PRs

PR	Title	Status	Checks
#10	ci: add test workflow	OPEN MERGEABLE	checks: none
#11	docs: fix package repository links	OPEN MERGEABLE	checks: none
#12	fix: add medea-eval entrypoint module	OPEN MERGEABLE	checks: none
#13	docs: add credential-light smoke test	OPEN MERGEABLE	checks: none
#14	docs: describe immune response data format	OPEN MERGEABLE	checks: none
#15	docs: clarify TargetID data inputs	OPEN MERGEABLE	checks: none

Open Issues

• #9 Question: Immune Therapy task
• #5 code for single-cell disease atlas processing

Local Test

Tests not run. Refresh with --run-tests.

Recent Commits

3f72428 (HEAD -> docs/targetid-data-inputs, fork/docs/targetid-data-inputs) docs: clarify TargetID data inputs
d44d1d4 (origin/main, origin/HEAD, fork/main, fork/HEAD, main) minor fix
90e18a3 Merge pull request #6 from felipeyanez/fix/get-efo-id-ols-wrong-ontology
92f302e Merge branch 'main' into fix/get-efo-id-ols-wrong-ontology
d7a9c50 Add more reasoning model support + Fix several minor issues

TxAgent

mims-harvard/TxAgent

Branch: docs/rag-reliability-checklist · HEAD: 6f2e182 · origin/main: 8c24ad9

clean working tree

Tracked PRs

PR	Title	Status	Checks
#19	ci: add compile workflow	OPEN MERGEABLE	checks: none
#20	fix: define token overflow dependencies	OPEN MERGEABLE	checks: none
#21	fix: correct assistant role typo	OPEN MERGEABLE	checks: none
#22	fix: defer heavy TxAgent runtime imports	OPEN MERGEABLE	checks: none
#23	docs: explain ToolRAG embedding cache	OPEN MERGEABLE	checks: none
#24	docs: clarify multi-GPU runtime setup	OPEN MERGEABLE	checks: none
#25	docs: add RAG reliability checklist	OPEN MERGEABLE	checks: none

Open Issues

• #18 Splitting model across 2 GPUs
• #16 Add a small RAG reliability checklist (docs only, no code changes)
• #15 Multiple function calls not implemented for 'llama' format.
• #14 Error: name "torch“, "logger" is not defined
• #13 Is there a manual evaluation of dataset generation?
• #12 Project requestments
• #11 TXAGENT-INSTRUCT dataset release plan
• #10 No such file or directory: 'ToolRAG-T1-GTE-Qwen2-1.5Btool_embedding[md value for tools].pt'
• #9 Will the benchmark data be open source?
• #8 Tokenizer

Local Test

Tests not run. Refresh with --run-tests.

Recent Commits

6f2e182 (HEAD -> docs/rag-reliability-checklist, fork/docs/rag-reliability-checklist) docs: add RAG reliability checklist
8c24ad9 (origin/main, origin/HEAD, fork/main, fork/HEAD, main) v0.1.3:fix gradio issue and vllm version issue
9659e4f Merge branch 'main' of github.com:mims-harvard/TxAgent
418e3ff add data file
2311070 Update example to match with the latest vllm version.