agent · coding assistant

Gemini Code Assist

cloud.google.comevidence retrieved Aug 1, 2026view as JSON all coding assistant entries 28 alternatives

Indexed12 or more fields evidenced; not yet scored by a human against the rubric

24 fields evidenced · 31 with no public information. Every value below links to the document it came from and the date we checked it.

24 of them are quoted but not yet interpreted: we hold the source sentence and the date we read it, but nobody has written the answer to the question yet. Those rows show the quote and say so rather than repeating it back as an answer. Why these exist

2 of those are flagged for review: our own check found language in the cited source that may address the field. An editor has not adjudicated them yet.

Also appears in

AI agents that support the Model Context Protocol Open-source AI agents AI agents that state a human-in-the-loop control

Trust gap

How this is scored

These scores are automated. A 1 or 2 counts how many fields this entry answers with a citation — not what those answers say. A 0 means every field for that dimension is recorded as undisclosed and none of the pages we captured mentions it. No one has scored this entry against the rubric yet.

Runtime governance

2/3 auto

Are scope, spend, and authority enforced while the agent runs?

Scored from 2 evidenced field(s): runtime_governance, permission_scopes. Automated score; capped at 2 pending review.

Counts Runtime governance and Permission scopes, each cited below.

Audit trail

unscored

Is there a tamper-evident record of what it actually did?

Compliance

1/3 auto

Are compliance obligations attached per engagement?

Scored from 1 evidenced field(s): compliance_certs. Automated score; capped at 2 pending review.

Counts Compliance certifications, each cited below.

Outcome settlement

unscored

Does payment depend on a verified result?

Insurance & indemnity

1/3 auto

Can the deployment be insured, and is the customer indemnified?

Scored from 1 evidenced field(s): indemnification. Automated score; capped at 2 pending review.

Counts Indemnification, each cited below.

Assurance

Insurance available: no public information
cloud.google.com · checked Aug 1, 2026 · source did not state this
Insurance carriers: no public information
cloud.google.com · checked Aug 1, 2026 · source did not state this
Coverage limits: no public information
cloud.google.com · checked Aug 1, 2026 · source did not state this
Indemnification: not yet interpreted
“Google’s IP indemnification policy helps protect Gemini Code Assist licensed users from potential legal ramifications concerning copyright infringements.”
cloud.google.com · checked Aug 1, 2026
Liability cap: no public information
cloud.google.com · checked Aug 1, 2026 · source did not state this
Audit trail: no public information
cloud.google.com · checked Aug 1, 2026 · source did not state this
Tamper-evident log: no public information
cloud.google.com · checked Aug 1, 2026 · source did not state this
Explainability: not yet interpreted
“Gemini Code Assist provides source citation so that code suggestions are automatically flagged when directly quoting at length from a source”
cloud.google.com · checked Aug 1, 2026
Runtime governance: not yet interpreted
“Enterprise Access Controls with granular IAM permissions to help enterprises adopt AI assistance at scale without compromising on security and privacy.”
cloud.google.com · checked Aug 1, 2026
Permission scopes: not yet interpreted
“Enterprise Access Controls with granular IAM permissions”
cloud.google.com · checked Aug 1, 2026
Compliance certifications: not yet interpreted
“Gemini Code Assist has achieved multiple industry certifications such as SOC 1/2/3, ISO/IEC 27001 (Information Security Management), 27017 (Cloud Security), 27018 (Protection of PII), and 27701 (Privacy Information Management).”
cloud.google.com · checked Aug 1, 2026
Regulatory alignment: no public information
cloud.google.com · checked Aug 1, 2026 · source did not state this
Outcome-based pricing: no public information
cloud.google.com · checked Aug 1, 2026 · source did not state this
Settlement mechanism: no public information
cloud.google.com · checked Aug 1, 2026 · source did not state this
Dispute process: no public information
cloud.google.com · checked Aug 1, 2026 · source did not state this
SLA terms: no public information
cloud.google.com · checked Aug 1, 2026 · source did not state this
Evaluation coverage: no public information
cloud.google.com · checked Aug 1, 2026 · source did not state this

Agency

Autonomy level: no public information
cloud.google.com · checked Aug 1, 2026 · source did not state this
Human oversight: not yet interpreted
“Human in the Loop (HiTL) for needed oversight.”
cloud.google.com · checked Aug 1, 2026
Goal complexity: not yet interpreted
“agents capable of performing a wide range of tasks across the software development lifecycle. It expands the capabilities of simple command-response interactions inside chat, with support for multiple file edits, full project context, built-in tools, and integration with ecosystem tools using MCP”
cloud.google.com · checked Aug 1, 2026
Action space: not yet interpreted
“from code understanding and file manipulation to command execution and dynamic troubleshooting”
cloud.google.com · checked Aug 1, 2026
Operating environment: not yet interpreted
“in their IDE, and Google Cloud services including Firebase, Colab Enterprise (Vertex AI), Databases BigQuery, Apigee, and Application Integration”
cloud.google.com · checked Aug 1, 2026
Initiative: not yet interpreted
“agents capable of performing a wide range of tasks across the software development lifecycle”
cloud.google.com · checked Aug 1, 2026

Safety

Safety evaluations: no public information
cloud.google.com · checked Aug 1, 2026 · source did not state this
Red teaming: no public information
cloud.google.com · checked Aug 1, 2026 · source did not state this
Safety policy: not yet interpreted
“Our data governance policy helps ensure customer code, customers' inputs, as well as the recommendations generated will not be used to train any shared models nor used to develop any products.”
cloud.google.com · checked Aug 1, 2026
Usage restrictions: no public information
cloud.google.com · checked Aug 1, 2026 · source did not state this
Model or system card: no public information
cloud.google.com · checked Aug 1, 2026 · source did not state this
Incident reporting: no public information
cloud.google.com · checked Aug 1, 2026 · source did not state this
Third-party evaluations: no public information
cloud.google.com · checked Aug 1, 2026 · source did not state this
Data handling: not yet interpreted
“Our data governance policy helps ensure customer code, customers' inputs, as well as the recommendations generated will not be used to train any shared models nor used to develop any products. Customers control and own their data and IP.”
cloud.google.com · checked Aug 1, 2026

Practicality

Pricing model: not yet interpreted
“Pricing is based on per user per month licenses”
cloud.google.com · checked Aug 1, 2026
Price point: no public information
cloud.google.com · checked Aug 1, 2026 · source did not state this
Availability: not yet interpreted
“currently in preview”
cloud.google.com · checked Aug 1, 2026
Deployment options: not yet interpreted
“VPC Service Controls”
cloud.google.com · checked Aug 1, 2026
Integrations: no public information
Flagged for review — the source below contains language that may address this field. Not yet checked by an editor. How we check absences
cloud.google.com · checked Aug 1, 2026
Supported regions: no public information
cloud.google.com · checked Aug 1, 2026 · source did not state this
Support model: no public information
cloud.google.com · checked Aug 1, 2026 · source did not state this

Foundation models

Base models: not yet interpreted
“Gemini 3 is our most intelligent model”
cloud.google.com · checked Aug 1, 2026
Model provider: not yet interpreted
“Gemini Code Assist uses large language models (LLMs) from Google.”
cloud.google.com · checked Aug 1, 2026
Model swappable: no public information
cloud.google.com · checked Aug 1, 2026 · source did not state this
Open weights: no public information
cloud.google.com · checked Aug 1, 2026 · source did not state this
Fine-tuning: not yet interpreted
“The LLMs are fine-tuned with billions of lines of open source code, security data, and Google Cloud documentation and sample code.”
cloud.google.com · checked Aug 1, 2026
Context window: not yet interpreted
“a 1M token context window”
cloud.google.com · checked Aug 1, 2026

Ecosystem

Protocols supported: not yet interpreted
“integration with ecosystem tools using MCP”
cloud.google.com · checked Aug 1, 2026
Tool use: not yet interpreted
“built-in tools, and integration with ecosystem tools using MCP”
cloud.google.com · checked Aug 1, 2026
Multi-agent: no public information
cloud.google.com · checked Aug 1, 2026 · source did not state this
API access: no public information
cloud.google.com · checked Aug 1, 2026 · source did not state this
Open source: not yet interpreted
“Gemini CLI , an open source AI agent”
cloud.google.com · checked Aug 1, 2026
Marketplace presence: no public information
cloud.google.com · checked Aug 1, 2026 · source did not state this

Impact

User base: no public information
Flagged for review — the source below contains language that may address this field. Not yet checked by an editor. How we check absences
cloud.google.com · checked Aug 1, 2026
Deployment scale: not yet interpreted
“Gemini Code Assist Standard and Enterprise AI-assisted development for your business”
cloud.google.com · checked Aug 1, 2026
Target sectors: not yet interpreted
“How Renault Group is using Google’s software-defined vehicle industry solution leveraging Gemini Code Assist”
cloud.google.com · checked Aug 1, 2026
High-risk domains: no public information
cloud.google.com · checked Aug 1, 2026 · source did not state this
Documented incidents: no public information
cloud.google.com · checked Aug 1, 2026 · source did not state this

Compared with

Side-by-side on the fields where Gemini Code Assist and the other entry are both on record and actually differ.