o2r: #2664 comment + trust-convention writeup (the deliverable) #35

New issue

Closed

opened 2026-06-02 09:26:09 +00:00 by coilysiren · 2 comments

coilysiren commented

2026-06-02 09:26:09 +00:00

Owner

Source: mobile session, concluding message (parsed into o2r tickets).

The actual deliverable — the thing you carry into the meetings. The meetings are access. This is what you bring to them. Without it, the meetings are just lurking.

Two parts:

Five-concept o2r mapping — the o2r concepts mapped onto the semconv / SIG vocabulary.
Inter-agent + MCP-boundary trust angle — the new contribution. This is the novel part, not the mapping.

Deliverable form: a comment on #2664 plus a standalone convention writeup.

Sequencing: blocked by #40 (the A2A security-model question) — that answer decides whether this trust convention is novel or already covered by Agent Cards and auth schemes.

Migrated from coilysiren/inbox#18.

**Source:** mobile session, concluding message (parsed into o2r tickets). The actual deliverable — the thing you carry into the meetings. The meetings are access. This is what you bring to them. Without it, the meetings are just lurking. Two parts: - **Five-concept o2r mapping** — the o2r concepts mapped onto the semconv / SIG vocabulary. - **Inter-agent + MCP-boundary trust angle** — the new contribution. This is the novel part, not the mapping. **Deliverable form:** a comment on #2664 plus a standalone convention writeup. **Sequencing:** blocked by #40 (the A2A security-model question) — that answer decides whether this trust convention is novel or already covered by Agent Cards and auth schemes. --- _Migrated from coilysiren/inbox#18._

coilysiren referenced this issue

2026-06-02 09:28:15 +00:00

o2r: A2A security-model question (run first — blocks trust convention) #40

coilysiren referenced this issue

2026-06-02 20:14:06 +00:00

o2r: A2A security-model question (run first — blocks trust convention) #40

coilysiren commented

2026-06-02 20:19:10 +00:00

Author

Owner

Draft landed - unblocked by #40

#40 is answered (see its comment): as an auth/identity mechanism the trust convention is redundant against A2A signed AgentCards + MCP OAuth 2.1. The novelty survives only as an observability contribution, in the #2664 lane. This draft reframes accordingly: the deliverable is trust-boundary semantic conventions, not "a trust convention."

Both deliverable forms drafted:

Standalone convention writeup - docs/trust-boundary-semconv.md (DRAFT, committed to the branch for review). Part 1 is the five-concept mapping (Session / Agent / Task / Message-Action / Graph) onto #2664 vocab + A2A objects + existing o2r attributes - the credibility move. Part 2 is the novel gen_ai.authorization.* attribute group + the MCP-boundary span-link correlation - the contribution.
#2664 comment - drafted at /tmp/o2r-2664-comment-draft.md, not posted. Posting to the public OTel repo is external/reputational and is Kai's call.

Open decisions before this ships:

Firm up the gen_ai.authorization.* attribute names against current gen_ai.* registry conventions.
Decide whether to lead the #2664 comment with the mapping (credibility-first) or the gap (contribution-first).
#38 landscape map should render the two empty coverage cells this identifies (acknowledgment legibility, A2A→MCP authorization correlation).

## Draft landed - unblocked by #40 #40 is answered (see its comment): as an auth/identity mechanism the trust convention is redundant against A2A signed AgentCards + MCP OAuth 2.1. The novelty survives only as an **observability** contribution, in the #2664 lane. This draft reframes accordingly: the deliverable is **trust-boundary semantic conventions**, not "a trust convention." **Both deliverable forms drafted:** - **Standalone convention writeup** - `docs/trust-boundary-semconv.md` (DRAFT, committed to the branch for review). Part 1 is the five-concept mapping (Session / Agent / Task / Message-Action / Graph) onto #2664 vocab + A2A objects + existing o2r attributes - the credibility move. Part 2 is the novel `gen_ai.authorization.*` attribute group + the MCP-boundary span-link correlation - the contribution. - **#2664 comment** - drafted at `/tmp/o2r-2664-comment-draft.md`, **not posted**. Posting to the public OTel repo is external/reputational and is Kai's call. **Open decisions before this ships:** - Firm up the `gen_ai.authorization.*` attribute names against current `gen_ai.*` registry conventions. - Decide whether to lead the #2664 comment with the mapping (credibility-first) or the gap (contribution-first). - #38 landscape map should render the two empty coverage cells this identifies (acknowledgment legibility, A2A→MCP authorization correlation).

coilysiren referenced this issue

2026-06-02 20:23:28 +00:00

o2r: implement gen_ai.authorization.* trust-boundary attributes + A2A->MCP span link (impl of #35) #42

coilysiren added the

label

2026-06-04 08:16:41 +00:00

coilysiren referenced this issue

2026-06-04 18:02:28 +00:00

Land trust-boundary-semconv.md draft writeup (Part 2 of #35) #58

coilysiren referenced this issue from a commit

2026-06-04 18:03:00 +00:00

docs(trust-boundary): land the trust-boundary semconv writeup draft

coilysiren commented

2026-06-17 08:23:50 +00:00

Author

Owner

Backlog burndown 2026-06-17: closing low-priority (P3/P4) to bring the open count to a manageable level. Nothing lost — reopen if this resurfaces. Batch tag: burndown-2026-06.

Backlog burndown 2026-06-17: closing low-priority (P3/P4) to bring the open count to a manageable level. Nothing lost — reopen if this resurfaces. Batch tag: `burndown-2026-06`.

coilysiren

2026-06-17 08:23:50 +00:00