PRODUCT PROFILE · RANK 8 OF 8
Prompt Security review: AI agent discovery and protection inside SentinelOne
SHORT ANSWER
Prompt Security inventories AI tools and code assistants, maps agents and MCP servers and protects homegrown AI apps, built into the SentinelOne Singularity Platform. It reaches devices through a browser extension and a dedicated AI endpoint agent; per-action enforcement is not described in detail. It scores 5.45 and ranks 8 of 8, highest or tied highest on Public proof.
By The Agent Desk · Agentic Endpoint Compare · Published 2026-10-11
Desk research from public vendor pages, press releases and standards pages read on 27 September and 11 October 2026. Editorial assessment; we did not run the products.
Rank 8 of 8 among broader AI security platforms on our published weights
SentinelOne (deal announced 5 August 2025; completion reported in September 2025)
SentinelOne Singularity estates
What does Prompt Security control?
Inventories AI tools and code assistants, including shadow AI, maps agents and MCP servers, and protects homegrown AI apps, built into the SentinelOne Singularity Platform.
How does Prompt Security reach devices?
Two local sensors, a browser extension and a dedicated AI endpoint agent, built into the SentinelOne Singularity Platform.
Source: Prompt Security blog: browser and endpoint sensors (updated 16 August 2026) · Read 11 Oct 2026
Which verdicts can Prompt Security return?
- Block or denyDocumented
- Ask a personNot described
- Alert or warnNot described
- Mask or modifyDocumented
- Simulate firstNot described
- Remove or scale backNot described
"Prevents prompt injection and data leakage; its browser extension inspects and redacts prompts."
Source: SentinelOne: Prompt Security, securing AI · Read 27 Sep 2026
Source: Prompt Security blog: browser and endpoint sensors (updated 16 August 2026) · Read 11 Oct 2026
How does Prompt Security score on each criterion?
| Criterion | Weight | Score | Why |
|---|---|---|---|
| Stop before it runs | 25% | 6 / 10 Lowest in set | Governs what agents and MCP servers can do; per-tool-call enforcement is not detailed. Source: SentinelOne: Prompt Security, securing AI · Read 27 Sep 2026 |
| Verdict range | 15% | 3 / 10 | Documents 2 of six verdict types: Block or deny and Mask or modify. "Prevents prompt injection and data leakage; its browser extension inspects and redacts prompts." Source: SentinelOne: Prompt Security, securing AI · Read 27 Sep 2026 Source: Prompt Security blog: browser and endpoint sensors (updated 16 August 2026) · Read 11 Oct 2026 |
| Capability rules | 15% | 5 / 10 Lowest in set | Maps agents and MCP servers and finds shadow MCP servers; capability rules are not itemized. Source: SentinelOne: Prompt Security, securing AI · Read 27 Sep 2026 |
| Actor trail | 10% | 5 / 10 Tied lowest | Agent versus person attribution is not described on the pages we read. Source: SentinelOne: Prompt Security, securing AI · Read 27 Sep 2026 |
| Rollout path | 10% | 6 / 10 | Two resident sensors, a browser extension and a dedicated AI endpoint agent, inside the SentinelOne Singularity Platform. Source: Prompt Security blog: browser and endpoint sensors (updated 16 August 2026) · Read 11 Oct 2026 |
| Cloud and workspace reach | 10% | 6 / 10 | Covers homegrown AI apps from testing to runtime as well as employee and developer AI use. Source: SentinelOne: Prompt Security, securing AI · Read 27 Sep 2026 |
| Named agent coverage | 5% | 5 / 10 | Names 3 of the six agents we track: Claude Code, Cursor and GitHub Copilot. Source: SentinelOne: Prompt Security, securing AI · Read 27 Sep 2026 |
| Public proof | 10% | 8 / 10 Tied highest | Backed by SentinelOne, which announced the acquisition on 5 August 2025; pricing not published. Source: SentinelOne press release: agreement to acquire Prompt Security (5 August 2025) · Read 27 Sep 2026 |
Which agents does Prompt Security name?
- Claude CodeNamed
- CursorNamed
- GitHub CopilotNamed
- OpenAI CodexNot named
- WindsurfNot named
- Claude DesktopNot named
Prompt Security names 3 of the six agents we track.
Source: SentinelOne: Prompt Security, securing AI · Read 27 Sep 2026
What does it cost?
Not published
SentinelOne publishes no pricing for Prompt Security.
Source: SentinelOne: Prompt Security, securing AI · Read 27 Sep 2026
Where is Prompt Security strongest, and where does it fall short?
Strongest
- Public proof: 8 / 10 (Tied highest)
- Fits security teams that already run SentinelOne Singularity.
- Covers homegrown AI apps from testing to runtime as well as employee and developer AI use.
Falls short
- Stop before it runs: 6 / 10; highest Bay 9 / 10
- Verdict range: 3 / 10; highest Bay and Onyx Security 7 / 10
- Capability rules: 5 / 10; highest Bay 9 / 10
- Actor trail: 5 / 10; highest Bay 8 / 10
- Rollout path: 6 / 10; highest Bay 9 / 10
- Cloud and workspace reach: 6 / 10; highest Onyx Security, Zenity and Noma Security 9 / 10
- Named agent coverage: 5 / 10; highest Zenity 7 / 10
- Per-tool-call enforcement is not described in detail.
- Its device sensors are resident software: a browser extension and an endpoint agent.
Key facts
| Fact | Detail and source |
|---|---|
| Discovery | Inventories every AI tool and code assistant in use, including unsanctioned shadow AI; maps every agent and MCP server and discovers shadow MCP servers. Source: SentinelOne: Prompt Security, securing AI · Read 27 Sep 2026 |
| Code assistants | Protects developers using GitHub Copilot, Cursor and Claude Code. Source: SentinelOne: Prompt Security, securing AI · Read 27 Sep 2026 |
| Ownership | SentinelOne announced a definitive agreement on 5 August 2025. Its blog, updated 3 September 2025, is titled SentinelOne Acquires Prompt Security, and Investing.com (secondary) reported from SentinelOne's 8-K that the deal closed on 5 September 2025 for about $180 million in cash and stock. Source: SentinelOne blog: SentinelOne Acquires Prompt Security (updated 3 September 2025) · Read 11 Oct 2026 |
| Sensors | The browser extension inspects and redacts prompts and discovers shadow AI; the endpoint agent detects new AI agents, skills and connectors and applies runtime policies. Prompt Security argues that MDM or EDR script integrations alone are not enough. Source: Prompt Security blog: browser and endpoint sensors (updated 16 August 2026) · Read 11 Oct 2026 |
| Company news | SentinelOne says Latio's 2026 AI Security Market Report named it an AI Security Platform Leader (15 September 2026, vendor-stated). Source: SentinelOne press release: Latio 2026 AI Security Market Report (15 September 2026) · Read 27 Sep 2026 |
How does Prompt Security compare head to head?
| Pair | Result |
|---|---|
| Bay matched with Prompt Security (SentinelOne) | Bay 7.20, Prompt Security (SentinelOne) 5.45 |
| Neo matched with Prompt Security (SentinelOne) | Neo 6.85, Prompt Security (SentinelOne) 5.45 |
| Onyx Security matched with Prompt Security (SentinelOne) | Onyx Security 6.85, Prompt Security (SentinelOne) 5.45 |
| Prompt Security (SentinelOne) matched with Zenity | Prompt Security (SentinelOne) 5.45, Zenity 6.75 |
| Noma Security matched with Prompt Security (SentinelOne) | Noma Security 6.55, Prompt Security (SentinelOne) 5.45 |
| Lasso Security matched with Prompt Security (SentinelOne) | Lasso Security 6.40, Prompt Security (SentinelOne) 5.45 |
| CrowdStrike Falcon Guardian matched with Prompt Security (SentinelOne) | CrowdStrike Falcon Guardian 6.25, Prompt Security (SentinelOne) 5.45 |
| Harmonic Security matched with Prompt Security (SentinelOne) | Harmonic Security 5.90, Prompt Security (SentinelOne) 5.45 |
| Ent matched with Prompt Security (SentinelOne) | Ent 5.80, Prompt Security (SentinelOne) 5.45 |
| Koi (Palo Alto Networks) matched with Prompt Security (SentinelOne) | Koi (Palo Alto Networks) 5.75, Prompt Security (SentinelOne) 5.45 |
| Glow Security matched with Prompt Security (SentinelOne) | Glow Security 5.40, Prompt Security (SentinelOne) 5.45 |
| Bloom Security matched with Prompt Security (SentinelOne) | Bloom Security 4.65, Prompt Security (SentinelOne) 5.45 |
Frequently asked questions
Can Prompt Security (SentinelOne) stop an agent action before it runs?
Governs what agents and MCP servers can do; per-tool-call enforcement is not detailed. We score this 6 out of 10.
Which verdicts does Prompt Security (SentinelOne) document?
Prompt Security (SentinelOne) documents 2 of six verdict types: Block or deny and Mask or modify. "Prevents prompt injection and data leakage; its browser extension inspects and redacts prompts."
How much does Prompt Security (SentinelOne) cost?
Not published. SentinelOne publishes no pricing for Prompt Security.