Your organization runs five Copilot agents (code review, issue triage, release notes, dependency updates, and security scanning) all configured to use the same shared GitHub App installation token stored in a central secrets manager. A security audit identifies this as a risk. What is the primary security concern with this architecture?
Select an answer to reveal the explanation.
Short Explanation and Infographic
Sharing one token among five agents is like giving five contractors the same master key and the same entry code. If one contractor is compromised, you can't selectively revoke just their access—you have to change the locks for everyone. Token isolation means each agent gets its own credential, so a breach is contained.
Full explanation below image
Full Explanation
Token isolation is a security architecture principle that mirrors least privilege at the credential level. When five agents share a single token:
1. Blast radius of compromise: If the code review agent is exploited via prompt injection, the attacker now holds a token with the combined capabilities of all five agents—including security scanning privileges that can potentially read vulnerability data.
2. Inability to revoke selectively: If a token is compromised, revoking it disrupts all five agents simultaneously. With per-agent tokens, you can revoke the compromised agent's token while the other four continue operating.
3. Audit ambiguity: All five agents appear identical in the GitHub audit log—there is no way to attribute a specific action to a specific agent.
4. Scope creep prevention: Each agent's token can be scoped to exactly the repositories and permissions it needs. A shared token must accommodate the union of all five agents' requirements, which is invariably broader than any single agent needs.
The correct architecture is to provision a separate GitHub App installation (or separate fine-grained PAT) for each agent, scoped to only that agent's repositories and permissions. Option A (latency) is an operational concern unrelated to security. Option C is incorrect—GitHub App tokens can be retrieved and used by multiple processes; there is no technical prohibition, only a security antipattern. Option D misidentifies the SPOF risk: the central secrets manager being a single point of failure is a separate availability concern from the shared token security risk.