Study OAuth2 concepts
Jump & context
Week 6: MCP Proxy Hardening, LangGraph Checkpoints, OAuth2 Delegation & Inter-Agent Bus
The five hours
- 01
Concepts & system design
60 min· read · map the mechanismR22 — RFC 9700 OAuth 2.0 Security BCP
R23 — RFC 8693 OAuth 2.0 Token Exchange
Read the assigned sections; map the mechanism, threat model and trade-offs.
- 02
Project implementation
120 min· build · integrate · debugStudy OAuth 2.0 Security Best Current Practice (RFC 9700) and Token Exchange (RFC 8693):
sub(user subject) vs.act(agent actor) claims, narrow scopes,aud(resource audience) binding, short TTL, and revocation. Design the scope matrix forP3.Build, integrate, debug and improve the increment in the active project.
- 03
Security lab
60 min· attack your own systemThreat-model token leakage, confused-deputy privilege escalation, and token replay across tools.
Test the abuse case, failure path, authorization bypass or resource limit.
- 04
Verification & evidence
30 min· test · record evidenceD7identity & delegation architecture diagram and scope matrix.Add or run regression tests; record the 9-point evidence record.
- 05
Deliberate practice & Socratic review
30 min· explain aloud · update gapsAnswer the 8 Socratic questions without notes: What? Why? How? What if? Why not? How can I prove it? Where can it break? What have I learned?
Re-explain the hardest concept aloud, inspect one edge case, update the gap log.
- total5 h
Read today · inside the concepts block
R22 R23RFC 6749 core flows + RFC 9700 security best practices: narrow scopes, audience restriction, token leakage, and replay.
Delegated authorization patterns, sub (subject) vs. act (actor) claims for agent delegation.
Read only the named sections needed for today's task. Reading sits inside the 60-minute concepts block; it is not extra time.
Free resources for this day
5 curated- docsRFC Editor (IETF) · ~60 min readRFC 9700: Best Current Practice for OAuth 2.0 Security
Sections 2.2 and 2.3 mandate replay prevention, audience restriction and narrow access-token privilege, which is the justification for short TTLs plus `aud` binding in the scope matrix.
R22moved 2026-11-11 - videoSascha Preibisch · ~10 minOAuth 2.0 - Token Exchange
Explains RFC 8693 exchange of one token for another with narrowed scope and a different audience, the core move the identity broker performs per tool call.
R23checked 2026-11-11 - videoGravitee · ~10 minStop Impersonating Users: Secure AI Agent Delegation with RFC 8693
Applies token exchange to agentic workflows and shows why delegation (with an actor identity) is safer than the agent impersonating the user outright.
R23checked 2026-11-11 - articleSagara Gunathunga · 14 min readOAuth 2.0 Token Exchange (RFC 8693) in Practice
Walks a real exchange through an API gateway and explains that only the outermost `act` claim is trusted for policy while nested claims are the delegation history.
R23checked 2026-11-11 - docsRFC Editor (IETF) · ~45 min readRFC 8693: OAuth 2.0 Token Exchange
Section 4.1 defines the `sub` vs `act` split and the nesting rule for delegation chains, the exact claim model your P3 scope matrix has to encode.
R23moved 2026-11-11
Evidence record · 9-point standard
A lab is complete only when it produces evidence. Installing PyRIT, Garak, Presidio or TransformerLens is not completion. Saved on this device only.
Day notes
emptyThe two source formulations
verbatim · nothing summarisedProject & Deliverable Target
P3 | `D7` (Delegated Agent OAuth Broker) | S6
Core Implementation
Study OAuth 2.0 Security Best Current Practice (RFC 9700) and Token Exchange (RFC 8693): sub (user subject) vs. act (agent actor) claims, narrow scopes, aud (resource audience) binding, short TTL, and revocation. Design the scope matrix for P3.
Security Lab
Threat-model token leakage, confused-deputy privilege escalation, and token replay across tools.
Verification & Evidence
D7 identity & delegation architecture diagram and scope matrix.
Read Today (60m Concepts Block)
R22 (RFC 9700 OAuth 2.0 Security BCP), R23 (RFC 8693 OAuth 2.0 Token Exchange).
---