gram
The AI control plane for every agent in your enterprise, built around an enterprise-grade MCP gateway. Give each team one governed path to its MCP servers, provisioned by role from the IdP you already run, with security policy enforced on every action.
No outside reports on file
Gram is a high-value target because it controls all agent access; misconfiguration or a single breach could compromise your entire agent ecosystem.
Based on public adoption and transparency signals as of Oct 19, 2026. Our opinion, not a code audit or security guarantee.
What it can access
- Local filesNo local folder access notedLOW
- Command executionNo shell command runner notedLOW
- Network reachFetches arbitrary web URLsREVIEW
- CredentialsNo API token requiredLOW
- Package sourceGitHub (speakeasy-api/gram)PUBLIC
Inferred from this listing’s public package details and description. Check the project’s source before installing.
Install
No npm or PyPI package is listed. Install from source: github.com/speakeasy-api/gram ↗
About Gram
Gram is a management layer that controls how AI agents access MCP servers (tools that connect AI systems to enterprise systems). It enforces security rules, user roles, and approval policies on every action an agent takes, sitting between your agents and the services they connect to.
Who it’s for
Enterprise teams that run multiple AI agents and need centralized control over which agents can access which systems. It's designed for companies using role-based access control (permissions tied to job titles) and single sign-on systems.
Alternatives
- OpenAI's built-in tool permissions: Basic role controls within OpenAI's platform
- HashiCorp Vault: General-purpose secrets and access management (not agent-specific)
- Kong API Gateway: General API management that can proxy agent requests
- Custom authorization server: Writing your own permission layer tailored to your agents
Setup
Check these steps against the project’s own instructions before running anything.
Follow the project's README. The GitHub repository should have installation instructions for deploying the MCP gateway in your environment.
Why this grade
Signal-based grade (3/7 points): listed in several sources, public source code. This reflects adoption and transparency, not a code review.
Popularity and upkeep
- GitHub stars
- 275
- Listing last refreshed
- 2026-10-29
These are public popularity signals, not a maintenance check. Look at the repository’s commit history and open issues.
Security reports
No reports found for this package in our research (last checked 2026-10-29).
Risks to watch for
- Becomes a single point of failure: If Gram is compromised or misconfigured, attackers could control all agent access to every connected system
- Over-trusts role data: If your identity provider (the system that says who you are) is hacked or returns false roles, Gram enforces wrong permissions
- Proxy-layer interception: Attackers who breach Gram's network position see all agent requests and responses flowing through it, including sensitive API calls
- Policy bypass via misconfiguration: Security rules can be accidentally weakened during setup, giving agents more access than intended
- Logs may contain secrets: Audit trails of agent actions could accidentally record API keys or passwords passed through requests
How to lower the risk
Least-privilege permissions
- Grant agents only the specific MCP servers and actions they actually need
- Regularly audit and remove unused permissions
Network isolation
- Run Gram in a private network segment, not on the public internet
- Restrict which machines can reach Gram using firewalls
Monitoring
- Log all policy decisions and agent requests for review
- Set alerts if an agent suddenly accesses new systems
Red flags to watch for
- Agents requesting permissions they shouldn't need
- Unexpected changes to policy files or role assignments
- High volumes of failed access attempts (signs of probing)
Questions
How do I install Gram?
Gram does not publish a standalone npm or PyPI package. Clone or follow the setup instructions in its source repository at github.com/speakeasy-api/gram.
Are there security reports about Gram?
No reports found for this package in our research (last checked 2026-10-29). Gram currently holds Grade C (Moderate risk signals) based on public adoption and transparency signals. Key risk areas to watch include becomes a single point of failure, over-trusts role data, proxy-layer interception.
Is Gram actively maintained?
Gram has 275 GitHub stars. Our directory data was last refreshed on 2026-10-29. Check recent commit activity, release tags, and open issues at github.com/speakeasy-api/gram before pinning it in production.
What are alternatives to Gram?
Notable alternatives to Gram include OpenAI's built-in tool permissions (Basic role controls within OpenAI's platform), HashiCorp Vault (General-purpose secrets and access management (not agent-specific)), Kong API Gateway (General API management that can proxy agent requests), Custom authorization server (Writing your own permission layer tailored to your agents). You can also compare Gram side by side with other skills in the Related tools section on this page.
README badge and team notes
Maintain or use Gram? Show its live grade in your GitHub README, docs site, or run a check from the terminal before you install.
GitHub README Markdown:
[](https://agentarmorhq.com/skills/speakeasy-api-gram)HTML embed:
<a href="https://agentarmorhq.com/skills/speakeasy-api-gram"><img src="https://agentarmorhq.com/api/badge/speakeasy-api-gram" alt="AgentArmorHQ Risk Grade for Gram" height="20" /></a>Terminal check:
curl -fsSL https://agentarmorhq.com/api/cli | node --input-type=module - check "speakeasy-api-gram"Share with your security lead
Need sign-off to use gram? Copy this note for Slack, a ticket or a pull request.
### AgentArmorHQ Security Review Memo: gram
- **Listing URL:** https://agentarmorhq.com/skills/speakeasy-api-gram
- **Type:** AI agent skill
- **Risk Grade:** Grade C (Moderate risk signals)
- **Outside Security Advisories on File:** 0
- **Public Adoption:** 275 GitHub stars
#### Permission & Blast-Radius Signals
- **Local files:** No local folder access noted
- **Command execution:** No shell command runner noted
- **Network reach:** Fetches arbitrary web URLs
- **Credentials:** No API token required
- **Package source:** GitHub (speakeasy-api/gram)
#### Recommended Hardened Setup
- **Source Repo:** https://github.com/speakeasy-api/gram
- **Controls:** Pin an exact version in `.mcp.json`, scope access to a single workspace folder, and use a least-privilege read-only token where supported.
_Based on public adoption and transparency signals as of Oct 19, 2026. Our opinion, not a code audit or security guarantee._Key facts
- Type
- AI agent skill
- Publisher
- speakeasy-api
- Source code
- github.com/speakeasy-api/gram
- Est. context load
- ~88 tokens idle · ~450 tokens when active (on-demand skill)
- Popularity
- 275 GitHub stars
- Last updated
- 2026-10-29
- Last graded
- 2026-10-19
Community
Discussions for Gram
Ask questions or share configuration tips. Governed by our Community Rules & Terms.
No posts match this filter yet.
Be the first to ask a question or share a working tip.
Check before you install, in Claude
Your assistant answers Block, Review or No known issues for any package before it installs it.
Get this check inside Claude