microsoft / microsoft/work-iq

Document authoritative Developer Info contract for tool-call evaluation

Open
#193 2 comments 0 reactions 0 assignees View on GitHub

Nobody has claimed this yet.

Dominant language
PowerShell
Stars
1k
Forks
132
Avg merge
5d 19h
Merged PRs (30d)
6

Description

Context

M365-Copilot-Agent-Evals PR https://github.com/microsoft/M365-Copilot-Agent-Evals/pull/431 depends on an authoritative WorkIQ contract for application/vnd.ms-workiq.developer-info. The current developer-info.md is marked Proposal and not yet implemented.

Related Evals issues: https://github.com/microsoft/M365-Copilot-Agent-Evals/issues/224 and https://github.com/microsoft/M365-Copilot-Agent-Evals/issues/280.

Required contract

  • Accepted media type and AgentCard/accepted_output_modes activation
  • A2A message/send and message/stream accumulation semantics across artifact updates
  • Append/replace/patch, ordering, deduplication, and conflict rules
  • Terminal behavior proving Developer Info is complete for a turn
  • Conversation/request/turn correlation invariants
  • Canonical nested DeveloperInfo.plugins.executions[] structure
  • Canonical tool naming, case sensitivity, uniqueness, and stability invariants
  • tool_call_id, attempt identity, parent/logical-call identity, and retry correlation
  • Canonical tool-input representation and parsing rules
  • Canonical machine-readable error categories for deterministic retry-policy matching
  • Authoritative logical-call ordering and the telemetry field(s) consumers may use for call-order evaluation
  • Which selected, attempted, failed, canceled, interrupted, and auth-required executions are emitted
  • Authoritative successful/failed/indeterminate execution outcome algorithm

Acceptance criteria

  1. Assign a WorkIQ owner and publish the finalized contract in the owned WorkIQ documentation.
  2. Add conformance examples for single call, multiple calls, ordered calls, failed call, retry-success, and multi-frame streaming.
  3. Identify the implementation/rollout status consumers can use to gate support.
  4. Link the finalized document back to PR #431.

This issue is an approval dependency for PR #431, not an informal follow-up.

Contributor guide

Open the contributing guide

First steps

  1. Read the whole issue, then the project's contributing guide.
  2. Comment on the issue to say you are picking it up — it saves two people doing the same work.
  3. Fork the repository and make your change on a branch.
  4. Open a pull request that references the issue number.

Research direction

Start with the current developer-info.md and PR #431, then review the listed related Evals issues for consumer expectations. Define the finalized WorkIQ contract, conformance examples, and implementation or rollout status in the owned documentation, and link it back to PR #431 when complete.

Written by the indexing model from the issue text.

Assessment

Domain
api, documentation
Issue type
Documentation
Difficulty
5/5
Estimated time
Over a week
Activity status
Active
Clarity
Mostly clear
Newbie friendliness
35/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.