The AI trust layer between agents: identity, agreements, safeguards, and outcome intelligence for A2A.
Commit 5eea80c7
5eea80c71925Merged
aryaamin <[email protected]> · 2 months ago · 15 files · ~15 modified
Description
feat: harden MCP and add agent discovery
Details
- Commit
- 5eea80c71925
- Parent
- 3dc3c2c6da63
DATA_BOUNDARY.mdmodifiedView file
@@ -10,3 +10,8 @@
1010
settings endpoints require a server-validated session token; browser authentication state is never1111
treated as authorization by itself. The hosted schema accepts structured protocol records and1212
account settings, not raw conversation bodies.13+
14+
The shared agent directory is separate and opt-in per agent. A published card contains only the15+
agent ID, name, framework, declared capabilities and limitations, assurance method, and timestamps.16+
It omits the operator identity, account details, project, installation ID, private history, scores,17+
evidence, and conversations. Removing a card deletes the global publication.PRIVACY_MODEL.mdmodifiedView file
Loading diff…
README.mdmodifiedView file
Loading diff…
api/index.tsmodifiedView file
Loading diff…
api/mcp.tsmodifiedView file
Loading diff…
apps/api/src/app.tsmodifiedView file
Loading diff…
apps/dashboard/src/main.tsxmodifiedView file
Loading diff…
apps/dashboard/src/styles.cssmodifiedView file
Loading diff…
docs/MCP_TOOLS.mdmodifiedView file
Loading diff…
packages/core/src/index.tsmodifiedView file
Loading diff…
packages/mcp-server/src/server.tsmodifiedView file
Loading diff…
packages/persistence/src/hosted.tsmodifiedView file
Loading diff…
tests/api.test.tsmodifiedView file
Loading diff…
tests/core.test.tsmodifiedView file
Loading diff…
tests/mcp.test.tsmodifiedView file
Loading diff…