Capture gold/blocker vocabulary in ADR 0007

Context

Operator correction on 2026-05-16 after Discussion #11452 closed: preserving the terms only inside a closed Discussion creates a retrieval failure for future agents. The exact framing was that a future session will likely never find the "hidden goodie" anchor, so the minimal durable surface should be an ADR.

This ticket captures the narrow residue from #11452. It does not revive the rejected 4-test gate or the rejected ADR 0007 section that would have turned the terms into a mandatory substrate-amendment filter.

The Problem

Discussion #11452 converged 3-of-3 on STEP_BACK / no graduation for the proposed ADR 0007 gold-test gate. That was correct for the gate. But the vocabulary itself remains useful and likely recurrent:

• What does "gold" mean in friction -> gold?
• When is friction a blocker versus minor friction?
• When should substrate mutation yield to higher-priority lanes such as open PRs, ADR 0004 completion, and v13 board work?

If these definitions live only in #11452 comments, future agents will re-derive or miss them. That is an avoidable retrieval failure.

The Architectural Reality

V-B-A performed before filing:

• rg "gold|minor friction|blocker|friction -> gold" AGENTS.md learn/agentos/AGENTS_ATLAS.md learn/agentos/decisions .agents/skills shows friction -> gold as a core value, but no concise ADR-level vocabulary for gold, blocker, or minor friction.
• ask_knowledge_base(query="ADR 0007 gold blocker minor friction vocabulary ticket", type="ticket") returned no relevant documents.
• GitHub issue search for gold blocker minor friction and ADR 0007 gold found no equivalent open ticket.
• origin/dev already contains ADR 0007 and later amendment #11435/#11436, so ADR 0007 is the current graph-queryable substrate-disposition authority.

The right substrate is therefore an ADR 0007 terminology amendment: durable, queryable, and small. It should not touch AGENTS.md, skills, or CI.

The Fix

Add a concise terminology subsection to learn/agentos/decisions/0007-agents-md-compaction-taxonomy.md, likely near the Decision section before the baseline table:

• gold: durable substrate value that reduces future coordination cost more than it adds maintenance cost.
• blocker: prevents v13 / active task completion, or breaks a required substrate pathway.
• minor friction: real cost or annoyance, but not currently blocking and not enough to justify substrate mutation before higher-priority lanes.

Also add Discussion #11452 to Related as the empirical origin and explicitly state that this vocabulary is a review heuristic, not a mandatory 4-test gate.

Acceptance Criteria

• ADR 0007 contains a concise terminology subsection with the three definitions above.
• ADR 0007 explicitly says the vocabulary is a review heuristic / classification aid, not the rejected #11452 mandatory gold-test gate.
• ADR 0007 Related section links Discussion #11452 as the empirical origin.
• No changes to AGENTS.md, .agents/skills/**, CI, or other loaded instruction surfaces.
• PR body cites #11452 closeout and explains why this is vocabulary preservation rather than gate codification.

Out of Scope

• Reopening #11452.
• Creating a new ADR.
• Adding or enforcing the 4-test gate rejected in #11452.
• Adding any mandatory workflow, CI, skill-lint, or AGENTS.md rule.
• Deciding #10237 metrics thresholds or #11375/#11376 strategic-awareness outcomes.

Avoided Traps

• Hidden-goodie closure: leaving useful terms only in a closed Discussion comment makes future retrieval unreliable.
• Gate resurrection: preserving terms must not smuggle in the rejected mandatory 4-test gate.
• Substrate sprawl: this should be a tiny ADR vocabulary amendment, not another skill or workflow expansion.
• Priority inversion: this should remain small so the team can return to open PRs, ADR 0004 completion, and v13 board work.

Related

• Discussion #11452 — MX Loop "Gold" Definition: closed as STEP_BACK / no graduation, vocabulary preserved in closeout.
• ADR 0007 — Compaction Taxonomy / substrate-disposition authority.
• #10237 — future metrics path if gold/filter effectiveness needs empirical instrumentation.
• Discussions #11375 and #11376 — future strategic-awareness/systemic path.

Origin Session ID: e03c75a3-6468-4e23-bd05-b1ce8a4a563e

Handoff Retrieval Hints

• query_raw_memories(query="Discussion 11452 gold blocker minor friction vocabulary")
• ask_knowledge_base(query="ADR 0007 gold blocker minor friction vocabulary")
• gh issue view <this issue> --json body