Propose five Mission Contract enhancements from implementation experience#10
Open
nickgamb-strata wants to merge 2 commits into
Open
Conversation
…perience Based on our implementation of intent-bound authorization for MCP tool governance, these proposals address gaps in the Mission Contract spec: 1. Tamper-evident mission log via hash chaining (audit entries) 2. Mission template discovery (PS endpoint for structured templates) 3. Mission constraint budgets (max_tool_calls, timeout_seconds) 4. Tool ordering constraints (required_predecessors, prohibited_sequences) 5. Denial category classification (authorization/intent/trajectory) All proposals are OPTIONAL extensions that complement the existing design. Reference implementation available upon request.
65845ee to
22f2a11
Compare
This was referenced Apr 16, 2026
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary
Based on our implementation of AAuth Mission Contracts for MCP (Model Context Protocol) tool governance, we identified five optional extensions that address gaps in the current Mission spec (Section 8). All proposals are backward-compatible and complement the existing design.
We originally built a "Desired Outcome Contract" (DOC) protocol for intent-bound authorization. After evaluating both approaches, we aligned with AAuth and migrated our POC to Mission Contracts. These proposals capture the novel concepts from DOC that proved valuable in practice.
Proposed Extensions
Tamper-evident mission log via hash chaining — Each audit entry chains to the previous via
SHA-256(prev_hash | action | params | result), seeded from the mission's s256. Addresses the lack of integrity mechanisms in the mission log (Section 8.3).Mission template discovery — Optional PS endpoint publishing structured mission templates that help agents compose better proposals. Reduces clarification rounds for first-time interactions.
Mission constraint budgets —
max_tool_callsandtimeout_secondsfields in the mission blob providing hard limits on agent execution, complementing contextual PS governance.Tool ordering constraints —
required_predecessorsandprohibited_sequencesencoding domain safety rules (e.g., must read account before modifying it).Denial category classification —
denial_categoryfield on permission responses distinguishingauthorization_denied,intent_denied, andtrajectory_deniedfor richer observability.Reference Implementation
All five proposals are demonstrated in a working POC with a Person Server (Go), MCP Auth Proxy (Python), and Enterprise Ledger (banking MCP server with tools of varying sensitivity). The POC implements the complete AAuth mission flow: template discovery → mission proposal → HITL approval with s256 computation → per-call permission for non-pre-approved tools → hash-chained audit logging. Reference implementation available upon request.
Test plan