Who it is for
A framework-agnostic runtime for continuously improving agents
pyproject.toml:12
PUBLIC REPORT · ADDED BY METAL MANTRA / AGENT SIGNAL v0.2
reef.
Scanned repository says: A framework-agnostic runtime for continuously improving agents
Metal Mantra scanned this public repository on its own initiative; the maintainer did not submit it. Maintainers: scan the repository yourself to take this report over, or contact us.
Inspect GitHub source0community interest
Votes show community interest, not safety. They never change this score.
64/100CCCNot security certification
Critical findings need attention.
Critical finding · details withheld for 30 days
Inspect the findingsRuntime safety, real-world agent capability, fitness for your use case or independent certification.
Scan scope200 / 611Partial eligible-file selection · not the entire repository
Recorded findings13Heuristic matches requiring human review
SnapshotOct 12, 2026UTC · engine 0.4.0
DNS at scan timeUnverifiedDomain control is not security certification
What the source
actually tells us.
Structured, cited information from the scanned archive. Repository-authored claims are labeled and never treated as verified capabilities.
200 / 611Partial selection
ReadRepository-authored information
2LICENSE, pyproject.toml
What it can do
No explicit capability or integration was extracted; that does not mean none exists.
What adoption takes
Reef's artifact and checkpoint functionality requires the git-lfs system
README.md:83Manifest declares Python >=3.12 as a requirement.
pyproject.toml:16Declares a dependency on aiohttp.
pyproject.toml:25Declares a dependency on alembic.
pyproject.toml:26Declares a dependency on huggingface_hub.
pyproject.toml:27Declares a dependency on pyyaml.
pyproject.toml:28Declares a dependency on reef-client.
pyproject.toml:29
Limits and terms
A license or copying file is present; its terms have not been reviewed.
LICENSE:1
Observed technical context
TypeScript · 2 selected filesPython · 198 selected filesTool-calling reference · static only
These are references in selected source files, not proof that a feature works at runtime.
Read from commit c996015 at Oct 12, 2026, 4:51 AM UTC. Source links open that exact commit. Repository statements are unreviewed and do not affect the score.
02 / FINDINGS FIRSTWhat needs
What needs
a closer look.
Critical details are withheld publicly for the first 30 days after a scan. A matched pattern is a lead for investigation, not a confirmed vulnerability.
Dependencies: 237 exact runtime versions from uv.lock checked against OSV.dev on Oct 12, 2026; 5 with known advisories. Reachability is not checked.
13 findings
CRITICALCritical finding · details withheld for 30 days
Critical finding · details withheld for 30 days
Details are withheld from public view for 30 days after the scan. The owner can see them in full.
Review the private assessment and address the finding.
CRITICALCritical finding · details withheld for 30 days
Critical finding · details withheld for 30 days
Details are withheld from public view for 30 days after the scan. The owner can see them in full.
Review the private assessment and address the finding.
CRITICALCritical finding · details withheld for 30 days
Critical finding · details withheld for 30 days
Details are withheld from public view for 30 days after the scan. The owner can see them in full.
Review the private assessment and address the finding.
MAJORUnsafe deserialization
Unsafe deserialization
The code loads data with a format that can run code while it is read, such as pickle or unrestricted YAML. A poisoned file, cache entry or downloaded model then runs with the agent's permissions.
Use a data-only format (JSON, safetensors, yaml.safe_load) for anything that is not fully trusted.
MAJORUnsafe deserialization
Unsafe deserialization
The code loads data with a format that can run code while it is read, such as pickle or unrestricted YAML. A poisoned file, cache entry or downloaded model then runs with the agent's permissions.
Use a data-only format (JSON, safetensors, yaml.safe_load) for anything that is not fully trusted.
MAJORRemote model code trusted
Remote model code trusted
The code loads a model with trust_remote_code enabled, which downloads and runs Python from the model's repository. Whoever controls that repository controls code running inside the agent.
Leave trust_remote_code off, or pin the model to a reviewed revision hash.
MAJORDomain ownership not verified
Domain ownership not verified
Metal Mantra could not confirm that the owner controls the stated domain. The domain may belong to someone else.
Verify the domain using the Metal Mantra DNS TXT challenge.
MAJORDependency with a high-severity advisory
Dependency with a high-severity advisory
transformers 5.5.4 (PyPI) has known advisories: GHSA-x9r9-c232-4q39, GHSA-xrqw-3rrv-vx5w, PYSEC-2026-3929, PYSEC-2026-4174 (highest severity high). Hugging Face Transformers downloads custom generation code before trust consent
Upgrade to a fixed version listed in the advisory and regenerate the lockfile.
MINORPrivacy Policy link not confirmed
Privacy Policy link not confirmed
No privacy policy link was found on the stated homepage. Buyers cannot see how the vendor handles their data.
Expose a clear Privacy Policy link on the official homepage.
MINORTerms link not confirmed
Terms link not confirmed
No terms link was found on the stated homepage. Buyers cannot see the conditions that apply to using the service.
Expose a clear Terms link on the official homepage.
MINORDependency with a known advisory
Dependency with a known advisory
accelerate 1.14.0 (PyPI) has known advisories: GHSA-4j2p-28q2-5m79, PYSEC-2026-3804 (highest severity moderate). Accelerate path traversal and denial of service via sharded checkpoint weight_map entries
Upgrade to a fixed version listed in the advisory and regenerate the lockfile.
MINORDependency with a known advisory
Dependency with a known advisory
mako 1.4.1 (PyPI) has a known advisory: GHSA-5639-2j2p-m4mx (highest severity moderate). Mako: Path traversal via drive-letter URI on Windows in TemplateLookup
Upgrade to a fixed version listed in the advisory and regenerate the lockfile.
MINORDependency with a known advisory
Dependency with a known advisory
multidict 6.7.1 (PyPI) has a known advisory: GHSA-54p9-h82j-f925 (highest severity moderate). Multidict: Reference leak in CIMultiDict/MultiDict items-view union and subtraction
Upgrade to a fixed version listed in the advisory and regenerate the lockfile.
The boundary
is part of the result.
- What was scanned
- 200 of 611 eligible files · Partial selection.
Supported static source patterns and domain checks; no runtime execution. - What remains unknown
- Unscanned files, external dependencies, production configuration, actual agent behavior and evidence beyond the configured checks.
- Source location
- Links open the current repository HEAD. Source may have changed since this snapshot.
- Something looks wrong?
- Tell us if a finding, scope or claim on this report needs a correction. Report a problem with this report
- Entity trust · 15% of v0.2
- Includes DNS ownership, HTTPS and policy evidence. Verification is domain control, not a security certificate.DNS: unverifiedHTTPS: availablePrivacy: not confirmedTerms: not confirmed
Open the
calculation.
Each pillar starts at 100. Rule deductions reduce its score; weights determine the overall contribution.
Security & privacy
0/100
Guardrails & loop control
100/100
Schema & tooling
100/100
Token & cost efficiency
100/100
Entity trust
25/100
Did reef work for people?
No user reports yet.
Each report is one signed-in GitHub account saying what happened. We do not verify that the account ran the agent, and reports never change the automated score or grade. For a security problem, use the disclosure page instead of a public note.
SHARE THE EVIDENCEA signal.
A signal.
Not a seal.
Share the full report so its scope and limits travel with the score.
