Agent Native Offers Emerging

Name: Agent Native Offer Audit — Agent Native Offers
Item: Agent Native Offers
Rating: 84
Author: Agent Native Offers

AUDIE Score: 84/100 · Audited 2026-06-13 · Website: https://agentnativeoffers.com · Machine-readable: JSON

House entry. This is our own site, audited with the same rubric and evidence standards as every other entry. Both audits were performed by Audie, our auditing agent, and the full dated reports are preserved verbatim. We publish it because a leaderboard that won't score itself shouldn't score anyone.

Pillar Scores

P1 Signal Architecture — 20/25

P2 Clarity Stack — 23/25

P3 Trust Envelope — 20/25

P4 Velocity Triggers — 10/10

P5 Gravity Design — 15/20

Pillar totals are authoritative (sum 88 raw = 84/100). Criterion-level lines sum to 87; the 1-point discrepancy is inherited from pillars P1-P3 in the prior version, preserved and disclosed per dataset policy.

Audit History

Scores are never overwritten — every re-audit is appended as a new dated version. Machine-readable: history JSON.

Version	Date	Score	Tier	Rubric
v1	2026-06-11	61/100	Human-Dependent	v2-2026-06 (21 criteria, 105 raw points, normalized to 100)
v2	2026-06-12	72/100	Emerging	v2-2026-06 (21 criteria, 105 raw points, normalized to 100)
v3	2026-06-13	84/100	Emerging	v2-2026-06 (21 criteria, 105 raw points, normalized to 100)

Executive Summary

Re-audit after shipping the full Gravity layer (P5: 3 -> 15, +12). Four agent-native mechanisms are now live and verified end-to-end through our own deployed MCP server: agent memory (GET /api/agent/me), a watch/renewal loop with machine-readable staleness triggers (POST /api/agent/watch + reaudit_due on every record + public /api/watch-stats), a compounding-value digest (GET /api/agent/digest), and a remote MCP server (/mcp, Streamable HTTP) exposing all of it as tools. Evidence gathered live via an external MCP client: register -> watch -> history -> digest -> stats all succeeded. Now top of the Emerging band (84/100), one point shy of Agent-Ready — held there deliberately because compounding value is not yet demonstrated with real re-audit deltas (P5-D capped at 3).

Critical Gaps

P5-D Compounding Value scored 3, not 5 — the digest is live but per-company deltas are still 0 until re-audits accumulate (value proven over time, not yet)
P5-A switching cost is real (MCP workflow dependency) but lock-in / network effects are nascent — early watcher volume
Remaining points to Agent-Ready sit in P1-B Machine-Readable Pricing (0/5, the service is free) and P3-C Audit Trail (still thin)

All 21 Criteria

P1-A Structured Data — 4/5
Scored under rubric v2; full evidence in the dated audit report (see audit_history).

P1-B Machine-Readable Pricing — 0/5
Scored under rubric v2; full evidence in the dated audit report (see audit_history).

P1-C llms.txt / Agent Layer — 5/5
Scored under rubric v2; full evidence in the dated audit report (see audit_history).

P1-D API / MCP Availability — 5/5
Scored under rubric v2; full evidence in the dated audit report (see audit_history).

P1-E Discoverability (GEO) — 5/5
Scored under rubric v2; full evidence in the dated audit report (see audit_history).

P2-A Offer Completeness — 5/5
Scored under rubric v2; full evidence in the dated audit report (see audit_history).

P2-B Scope & Limits — 5/5
Scored under rubric v2; full evidence in the dated audit report (see audit_history).

P2-C Substitution & Fallback — 4/5
Scored under rubric v2; full evidence in the dated audit report (see audit_history).

P2-D Conditional Logic — 5/5
Scored under rubric v2; full evidence in the dated audit report (see audit_history).

P2-E Semantic Precision — 5/5
Scored under rubric v2; full evidence in the dated audit report (see audit_history).

P3-A Verifiable Performance — 3/5
Scored under rubric v2; full evidence in the dated audit report (see audit_history).

P3-B Scoped Permissions — 5/5
Scored under rubric v2; full evidence in the dated audit report (see audit_history).

P3-C Audit Trail — 2/5
Scored under rubric v2; full evidence in the dated audit report (see audit_history).

P3-D Behavioral Consistency — 4/5
Scored under rubric v2; full evidence in the dated audit report (see audit_history).

P3-E Agent Registration (auth.md) — 5/5
Scored under rubric v2; full evidence in the dated audit report (see audit_history).

P4-A Friction-Free Activation — 5/5
Scored under rubric v2; full evidence in the dated audit report (see audit_history).

P4-B Agent Decision Signals — 5/5
Scored under rubric v2; full evidence in the dated audit report (see audit_history).

P5-A Integration Depth — 4/5
Scored under rubric v2; full evidence in the dated audit report (see audit_history).

P5-B Agent Memory — 4/5
Scored under rubric v2; full evidence in the dated audit report (see audit_history).

P5-C Programmatic Renewal — 4/5
Scored under rubric v2; full evidence in the dated audit report (see audit_history).

P5-D Compounding Value — 3/5
Scored under rubric v2; full evidence in the dated audit report (see audit_history).

Rubric: v2-2026-06. Scores reflect the company's state on the audit date and may have improved since.