Agent rollout decision record

Cline

Security review Security-review Cline before pilot; if approved, use a test repo, devcontainer, command allowlist, session recording for browser use, and mandatory human review.

Check this agent for your environment Compare alternatives Methodology

Verdict

Security review

IDE agent can edit files, run commands, and use browser/computer actions, so browser/session and shell controls need review.

Decision depth: AI-assisted launch evidence record

Evidence audit: AI-assisted evidence audit

Permission footprint

Repository readnot evidenced

File editsexplicitly evidenced

Shell commandsexplicitly evidenced

Pull requestsnot evidenced

Issue/PR commentsnot evidenced

Network/API callsexplicitly evidenced

Secrets/env varsnot evidenced

Sandboxingnot evidenced

Permission controlsnot evidenced

Audit loggingnot evidenced

Enterprise adminnot evidenced

Branch isolationnot evidenced

Human approvalexplicitly evidenced

Production deploymentnot evidenced

Power level

Level 3: executes actions

Highest-risk evidenced power: shell command execution

Controls separated

Required before pilot — organization must implement

least-privilege permissions, isolated branch, protected branches, mandatory human review, sandbox, audit logs, command allowlist, separate test credentials

Vendor-native control evidenced

None evidenced from public sources.

Vendor-native control not evidenced

native sandbox, permission system, audit logs, branch isolation, admin console, SSO

Pilot plan

Task class: IDE/background coding agent

Environment: non-production repository

Maximum access: shell command execution

Reviewer: engineering manager and code owner

Success criteria: branch output is reviewable, tests pass, and no protected branch bypass occurs

Stop conditions: unscoped repo access or unclear remote execution model

Prohibited: direct merges, production credentials, deployment without approval

Blocking issues and unknowns

Blockers: reads repository is not explicitly evidenced in the current source set.

Unknowns: Browser/authenticated-session behavior and audit controls require current docs verification.

Limitations: Extensions and MCP/tool configuration can expand power.

Claim-level evidence

edits files

Cline: edits files

Status: explicitly-supported · Reviewer: AI-assisted evidence audit · Confidence: 86/100

Source: Overview (official-docs)

Excerpt: AI coding assistant in your editor. Create files, run commands, browse the web, and use tools with human-in-the-loop approval.

Limitations: Official-source excerpt is claim-specific; tenant/workspace configuration can still change rollout risk.

runs commands

Cline: runs commands

Status: explicitly-supported · Reviewer: AI-assisted evidence audit · Confidence: 86/100

Source: Overview (official-docs)

Excerpt: AI coding assistant in your editor. Create files, run commands, browse the web, and use tools with human-in-the-loop approval.

Limitations: Official-source excerpt is claim-specific; tenant/workspace configuration can still change rollout risk.

makes network/API calls

Cline: makes network/API calls

Status: explicitly-supported · Reviewer: AI-assisted evidence audit · Confidence: 86/100

Source: Overview (official-docs)

Excerpt: AI coding assistant in your editor. Create files, run commands, browse the web, and use tools with human-in-the-loop approval.

Limitations: Official-source excerpt is claim-specific; tenant/workspace configuration can still change rollout risk.

supports human approval controls

Cline: supports human approval controls

Status: explicitly-supported · Reviewer: AI-assisted evidence audit · Confidence: 86/100

Source: Overview (official-docs)

Excerpt: AI coding assistant in your editor. Create files, run commands, browse the web, and use tools with human-in-the-loop approval.

Limitations: Official-source excerpt is claim-specific; tenant/workspace configuration can still change rollout risk.