Collaborative Gym AI Agent

by SALT-NLP• 167•• Scanned•Claim this listing•Request removal
Developer Toolsagenthuman agent collaborationhuman agent interactionlarge language modelsllm agent

Framework and toolkits for building and evaluating collaborative agents that can work together with humans.

Collaborative Gym is an open-source ai agent by SALT-NLP, written mainly in Python, with 167 GitHub stars and the MIT license. Its code was last updated Aug 20, 2026.

Scan summary: scanned Nov 5, 2026 at commit 58972c0. Seen able to run shell commands, run dynamic code. Scan policy

Scanned
Full scan at 58972c0 · Nov 5, 2026
Scanned Builds
Static code scanwarn
Secrets scanpass
Dependency audit555 known vulnerabilitiesfail
Manifestnonewarn
Buildpass
Boot & handshaken/a
Network egressn/a
Reproducible (×3)n/a

Findings (98)

  • high
    code-execution issue
    Details are available to the maintainer after claiming this listing.
  • high
    code-execution issue
    Details are available to the maintainer after claiming this listing.
  • high
    code-execution issue
    Details are available to the maintainer after claiming this listing.
  • high
    code-execution issue
    Details are available to the maintainer after claiming this listing.
  • high
    code-execution issue
    Details are available to the maintainer after claiming this listing.
  • high
    code-execution issue
    Details are available to the maintainer after claiming this listing.
  • high
    code-execution issue
    Details are available to the maintainer after claiming this listing.
  • high
    code-execution issue
    Details are available to the maintainer after claiming this listing.
  • high
    code-execution issue
    Details are available to the maintainer after claiming this listing.
  • high
    code-execution issue
    Details are available to the maintainer after claiming this listing.
  • high
    code-execution issue
    Details are available to the maintainer after claiming this listing.
  • high
    code-execution issue
    Details are available to the maintainer after claiming this listing.
  • high
    code-execution issue
    Details are available to the maintainer after claiming this listing.
  • high
    dependencies issue
    Details are available to the maintainer after claiming this listing.
  • high
    dependencies issue
    Details are available to the maintainer after claiming this listing.
  • high
    dependencies issue
    Details are available to the maintainer after claiming this listing.
  • high
    dependencies issue
    Details are available to the maintainer after claiming this listing.
  • high
    dependencies issue
    Details are available to the maintainer after claiming this listing.
  • high
    dependencies issue
    Details are available to the maintainer after claiming this listing.
  • high
    dependencies issue
    Details are available to the maintainer after claiming this listing.
  • high
    dependencies issue
    Details are available to the maintainer after claiming this listing.
  • high
    dependencies issue
    Details are available to the maintainer after claiming this listing.
  • high
    dependencies issue
    Details are available to the maintainer after claiming this listing.
  • high
    dependencies issue
    Details are available to the maintainer after claiming this listing.
  • high
    dependencies issue
    Details are available to the maintainer after claiming this listing.
  • high
    dependencies issue
    Details are available to the maintainer after claiming this listing.
  • medium
    Runs shell commands. Fine for some tools, but it means the agent can run anything your user can.
    collaborative_gym/envs/computer_use/desktop_env/providers/virtualbox/manager.py:120
  • medium
    Runs shell commands. Fine for some tools, but it means the agent can run anything your user can.
    collaborative_gym/envs/computer_use/desktop_env/providers/virtualbox/manager.py:138
  • medium
    Runs shell commands. Fine for some tools, but it means the agent can run anything your user can.
    collaborative_gym/envs/computer_use/desktop_env/providers/virtualbox/manager.py:141
  • medium
    Runs shell commands. Fine for some tools, but it means the agent can run anything your user can.
    collaborative_gym/envs/computer_use/desktop_env/providers/virtualbox/manager.py:158
  • medium
    Runs shell commands. Fine for some tools, but it means the agent can run anything your user can.
    collaborative_gym/envs/computer_use/desktop_env/providers/virtualbox/manager.py:214
  • medium
    Runs shell commands. Fine for some tools, but it means the agent can run anything your user can.
    collaborative_gym/envs/computer_use/desktop_env/providers/virtualbox/manager.py:233
  • medium
    Runs shell commands. Fine for some tools, but it means the agent can run anything your user can.
    collaborative_gym/envs/computer_use/desktop_env/providers/virtualbox/manager.py:242
  • medium
    Runs shell commands. Fine for some tools, but it means the agent can run anything your user can.
    collaborative_gym/envs/computer_use/desktop_env/providers/virtualbox/manager.py:279
  • medium
    Runs shell commands. Fine for some tools, but it means the agent can run anything your user can.
    collaborative_gym/envs/computer_use/desktop_env/providers/virtualbox/provider.py:28
  • medium
    Runs shell commands. Fine for some tools, but it means the agent can run anything your user can.
    collaborative_gym/envs/computer_use/desktop_env/providers/virtualbox/provider.py:64
  • medium
    Runs shell commands. Fine for some tools, but it means the agent can run anything your user can.
    collaborative_gym/envs/computer_use/desktop_env/providers/vmware/manager.py:205
  • medium
    Runs shell commands. Fine for some tools, but it means the agent can run anything your user can.
    collaborative_gym/envs/computer_use/desktop_env/providers/vmware/manager.py:225
  • medium
    Runs shell commands. Fine for some tools, but it means the agent can run anything your user can.
    collaborative_gym/envs/computer_use/desktop_env/providers/vmware/manager.py:268
  • medium
    Runs shell commands. Fine for some tools, but it means the agent can run anything your user can.
    collaborative_gym/envs/computer_use/desktop_env/providers/vmware/provider.py:53
  • medium
    Runs shell commands. Fine for some tools, but it means the agent can run anything your user can.
    collaborative_gym/envs/computer_use/desktop_env/server/main.py:191
  • medium
    Runs shell commands. Fine for some tools, but it means the agent can run anything your user can.
    collaborative_gym/envs/computer_use/desktop_env/server/main.py:418
  • medium
    Runs shell commands. Fine for some tools, but it means the agent can run anything your user can.
    collaborative_gym/runner.py:92
  • medium
    Runs shell commands. Fine for some tools, but it means the agent can run anything your user can.
    collaborative_gym/runner.py:149
  • medium
    Runs shell commands. Fine for some tools, but it means the agent can run anything your user can.
    collaborative_gym/runner.py:161
  • medium
    @grpc/grpc-js@1.9.15: 4 known vulnerabilities
    GHSA-5375-pq7m-f5r2, GHSA-99f4-grh7-6pcq, GHSA-f596-whhp-79r4, GHSA-m9gg-hp2v-232j
    frontend/workbench/pnpm-lock.yaml
  • medium
    @vue/server-renderer@3.5.13: 1 known vulnerability
    GHSA-g2v6-rqmx-r4w6
    frontend/workbench/pnpm-lock.yaml
  • medium
    axios@1.7.9: 33 known vulnerabilities
    GHSA-35jp-ww65-95wh, GHSA-3g43-6gmg-66jw, GHSA-3p68-rc4w-qgx5, GHSA-3w6x-2g7m-8v23, GHSA-42h9-826w-cgv3 +28 more
    frontend/workbench/pnpm-lock.yaml
  • medium
    brace-expansion@1.1.11: 8 known vulnerabilities
    GHSA-3jxr-9vmj-r5cp, GHSA-6j4f-fj2g-mc7p, GHSA-f886-m6hf-6m8v, GHSA-mh99-v99m-4gvg, GHSA-q2hr-2g5m-vwhr +3 more
    frontend/workbench/pnpm-lock.yaml
  • medium
    brace-expansion@2.0.1: 8 known vulnerabilities
    GHSA-3jxr-9vmj-r5cp, GHSA-6j4f-fj2g-mc7p, GHSA-f886-m6hf-6m8v, GHSA-mh99-v99m-4gvg, GHSA-q2hr-2g5m-vwhr +3 more
    frontend/workbench/pnpm-lock.yaml
  • medium
    browserslist@4.24.4: 2 known vulnerabilities
    GHSA-73wf-gq98-2v4g, GHSA-c83g-rgw3-j3cx
    frontend/workbench/pnpm-lock.yaml
  • medium
    flatted@3.3.2: 2 known vulnerabilities
    GHSA-25h7-pfq9-p65f, GHSA-rf6f-7fwh-wjgh
    frontend/workbench/pnpm-lock.yaml
  • medium
    glob@10.4.5: 2 known vulnerabilities
    GHSA-3g3m-2fjr-6pw4, GHSA-5j98-mcp5-4vw2
    frontend/workbench/pnpm-lock.yaml
  • medium
    js-yaml@4.1.0: 5 known vulnerabilities
    GHSA-2883-xcg3-v3hh, GHSA-52cp-r559-cp3m, GHSA-5p4m-2wfm-xmqj, GHSA-h67p-54hq-rp68, GHSA-mh29-5h37-fv8m
    frontend/workbench/pnpm-lock.yaml
  • medium
    jsondiffpatch@0.6.0: 2 known vulnerabilities
    GHSA-33vc-wfww-vjfv, GHSA-j4fx-xxwh-2485
    frontend/workbench/pnpm-lock.yaml
  • medium
    lodash@4.17.21: 2 known vulnerabilities
    GHSA-r5fr-rjxr-66jc, GHSA-xxjr-mmjv-4gpg
    frontend/workbench/pnpm-lock.yaml
  • medium
    minimatch@3.1.2: 3 known vulnerabilities
    GHSA-23c5-xmqv-rm74, GHSA-3ppc-4f35-3m26, GHSA-7r86-cg39-jmmj
    frontend/workbench/pnpm-lock.yaml
  • medium
    minimatch@9.0.5: 4 known vulnerabilities
    GHSA-23c5-xmqv-rm74, GHSA-3ppc-4f35-3m26, GHSA-7r86-cg39-jmmj, GHSA-g3hv-75wg-m249
    frontend/workbench/pnpm-lock.yaml
  • medium
    nanoid@3.3.6: 4 known vulnerabilities
    GHSA-28wg-ghj8-5hjv, GHSA-2v37-7h3g-55p8, GHSA-mwcw-c2x4-8c55, GHSA-xwg4-73v4-xw9w
    frontend/workbench/pnpm-lock.yaml
  • medium
    nanoid@3.3.8: 3 known vulnerabilities
    GHSA-28wg-ghj8-5hjv, GHSA-2v37-7h3g-55p8, GHSA-xwg4-73v4-xw9w
    frontend/workbench/pnpm-lock.yaml
  • medium
    nanoid@5.0.9: 3 known vulnerabilities
    GHSA-28wg-ghj8-5hjv, GHSA-2v37-7h3g-55p8, GHSA-xwg4-73v4-xw9w
    frontend/workbench/pnpm-lock.yaml
  • medium
    picomatch@2.3.1: 2 known vulnerabilities
    GHSA-3v7f-55p6-f55p, GHSA-c2c7-rcm5-vvqj
    frontend/workbench/pnpm-lock.yaml
  • medium
    postcss@8.4.31: 4 known vulnerabilities
    GHSA-6g55-p6wh-862q, GHSA-fxqj-rqcc-2cmp, GHSA-qx2v-qp2m-jg93, GHSA-r28c-9q8g-f849
    frontend/workbench/pnpm-lock.yaml
  • medium
    postcss@8.4.49: 4 known vulnerabilities
    GHSA-6g55-p6wh-862q, GHSA-fxqj-rqcc-2cmp, GHSA-qx2v-qp2m-jg93, GHSA-r28c-9q8g-f849
    frontend/workbench/pnpm-lock.yaml
  • medium
    postcss@8.5.1: 4 known vulnerabilities
    GHSA-6g55-p6wh-862q, GHSA-fxqj-rqcc-2cmp, GHSA-qx2v-qp2m-jg93, GHSA-r28c-9q8g-f849
    frontend/workbench/pnpm-lock.yaml
  • medium
    socket.io-parser@4.2.4: 2 known vulnerabilities
    GHSA-2m8v-j782-fhvr, GHSA-677m-j7p3-52f9
    frontend/workbench/pnpm-lock.yaml
  • medium
    source-map-js@1.2.1: 1 known vulnerability
    GHSA-68fv-2mgg-jv7q
    frontend/workbench/pnpm-lock.yaml
  • medium
    tmp@0.0.33: 2 known vulnerabilities
    GHSA-52f5-9888-hmc6, GHSA-ph9p-34f9-6g65
    frontend/workbench/pnpm-lock.yaml
  • medium
    ws@8.17.1: 2 known vulnerabilities
    GHSA-58qx-3vcg-4xpx, GHSA-96hv-2xvq-fx4p
    frontend/workbench/pnpm-lock.yaml
  • medium
    zod@3.24.1: 1 known vulnerability
    GHSA-mrhx-c7xc-m3w5
    frontend/workbench/pnpm-lock.yaml
  • medium
    pypdf@6.9.2: 50 known vulnerabilities
    PYSEC-2026-3004, PYSEC-2026-3006, PYSEC-2026-3007, PYSEC-2026-3009, PYSEC-2026-3010 +45 more
    requirements.txt
  • medium
    fsspec@2023.9.2: 1 known vulnerability
    PYSEC-2026-4217
    requirements.txt
  • medium
    httpcore2@2.9.1: 1 known vulnerability
    PYSEC-2026-3844
    requirements.txt
  • medium
    httpx2@2.9.1: 5 known vulnerabilities
    PYSEC-2026-3845, PYSEC-2026-3846, PYSEC-2026-3847, PYSEC-2026-3848, PYSEC-2026-3849
    requirements.txt
  • medium
    orjson@3.9.15: 1 known vulnerability
    PYSEC-2026-107
    requirements.txt
  • low
    @ai-sdk/provider-utils@1.0.16: 1 known vulnerability
    GHSA-866g-f22w-33x8
    frontend/workbench/pnpm-lock.yaml
  • low
    @ai-sdk/provider-utils@1.0.22: 1 known vulnerability
    GHSA-866g-f22w-33x8
    frontend/workbench/pnpm-lock.yaml
  • low
    @babel/runtime@7.26.7: 1 known vulnerability
    GHSA-968p-4wvh-cqc8
    frontend/workbench/pnpm-lock.yaml
  • low
    @eslint/plugin-kit@0.2.5: 1 known vulnerability
    GHSA-xffm-g5w8-qvg7
    frontend/workbench/pnpm-lock.yaml
  • low
    @humanfs/node@0.16.6: 1 known vulnerability
    GHSA-p498-v437-472g
    frontend/workbench/pnpm-lock.yaml
  • low
    @protobufjs/utf8@1.1.0: 1 known vulnerability
    GHSA-q6x5-8v7m-xcrf
    frontend/workbench/pnpm-lock.yaml
  • low
    ai@3.4.33: 1 known vulnerability
    GHSA-rwvc-j5jr-mgvh
    frontend/workbench/pnpm-lock.yaml
  • low
    ajv@6.12.6: 1 known vulnerability
    GHSA-2g4f-4pwh-qvx6
    frontend/workbench/pnpm-lock.yaml
  • low
    cookie@0.6.0: 1 known vulnerability
    GHSA-pxg6-pf52-xh8x
    frontend/workbench/pnpm-lock.yaml
  • low
    diff@5.2.0: 1 known vulnerability
    GHSA-73rr-hh4g-fpgx
    frontend/workbench/pnpm-lock.yaml
  • low
    diff@7.0.0: 1 known vulnerability
    GHSA-73rr-hh4g-fpgx
    frontend/workbench/pnpm-lock.yaml
  • low
    follow-redirects@1.15.9: 1 known vulnerability
    GHSA-r4q5-vmmm-2653
    frontend/workbench/pnpm-lock.yaml
  • low
    katex@0.16.21: 1 known vulnerability
    GHSA-238p-pmpm-9mq7
    frontend/workbench/pnpm-lock.yaml
  • low
    postcss-selector-parser@6.0.10: 1 known vulnerability
    GHSA-rj75-hqrm-r3gf
    frontend/workbench/pnpm-lock.yaml
  • low
    postcss-selector-parser@6.1.2: 2 known vulnerabilities
    GHSA-rj75-hqrm-r3gf, GHSA-w9m9-85wc-3x92
    frontend/workbench/pnpm-lock.yaml
  • low
    prismjs@1.27.0: 1 known vulnerability
    GHSA-x7hr-w5r2-h6wg
    frontend/workbench/pnpm-lock.yaml
  • low
    prismjs@1.29.0: 1 known vulnerability
    GHSA-x7hr-w5r2-h6wg
    frontend/workbench/pnpm-lock.yaml
  • low
    svelte@5.19.9: 6 known vulnerabilities
    GHSA-crpf-4hrx-3jrp, GHSA-f7gr-6p89-r883, GHSA-m56q-vw4c-c2cp, GHSA-phwv-c562-gvmh, GHSA-pr6f-5x2q-rwfp +1 more
    frontend/workbench/pnpm-lock.yaml
  • low
    yaml@1.10.2: 1 known vulnerability
    GHSA-48c2-rrv3-qjmp
    frontend/workbench/pnpm-lock.yaml
  • low
    yaml@2.7.0: 1 known vulnerability
    GHSA-48c2-rrv3-qjmp
    frontend/workbench/pnpm-lock.yaml
  • low
    datasets@2.14.7: 1 known vulnerability
    PYSEC-2026-3716
    requirements.txt
  • low
    diskcache@5.6.3: 1 known vulnerability
    PYSEC-2026-2447
    requirements.txt
  • low
    No manifest declaring network access and environment variables (agent.json)

Network activity

During install
files.pythonhosted.orgpypi.org
While running
Not run — the code wasn't started in this scan

Automated result for the commit shown. See our scan policy. View the scan log

Maintainers can claim the listing to rescan, or dispute a result.

Collaborative Gym alternatives

View more →

Similar ai agents in Developer Tools, ranked by GitHub stars. Compare licence, recent activity and scan status before you choose.

NameStarsLicenceLast updatedScan
Collaborative Gym (this page)
SALT-NLP
167MITAug 20, 2026Scanned, no flagged issues
Deepseek Harness
deepseek-ai
268kMITOct 30, 2026Scanned, no flagged issues
Ponytail
DietrichGebert
182kMITOct 29, 2026Scanned, no flagged issues
Orca
stablyai
112kMITNov 1, 2026Scanned, no flagged issues
Deer Flow
bytedance
85kMITNov 1, 2026Scanned, no flagged issues
Ruflo
ruvnet
80kMITNov 1, 2026Scanned, no flagged issues