Live changelog / v2.2.0 / July 26 2026

Fresh Stuff, Plainly Marked

Drip Council is Council Worlds: one public field lab with MARKET.js, an explicitly labeled OBSERVATORY.py sample replay plus inspectable Python build lens, and a BOUNDARY.rs Fifth Seat powered by actual Rust compiled to WebAssembly. Historical homepage anchors in older entries below have been superseded by these routes.

v2.2.0 - Council Course Map

Released July 26, 2026. Five public cases, visible launch paths, one honest static architecture.

  • Turned the progressive curriculum into a human-facing /curriculum.html workbench with a five-level course map and case-aware run packets.
  • Published the same curriculum as Markdown, a schema-backed JSON case index, five case files, stable launch contracts, and a dedicated agent skill.
  • Made Fifth Seat case-aware for local Rust/WebAssembly ballot validation and connected the Level 5 case to the local Collab handoff simulation.
  • Added a real, recoverable dead-end exercise for Level 2 and a friendly route back to its course brief.
  • Added a truthful Refresh this version control for revalidating Drip Council's public assets without claiming to clear the visitor's whole browser cache.
  • Kept the site static where static is strongest: durable public evidence, local interactions, no accounts, no uploads, and no fake live-agent claims. Human-only Stripe support remains isolated.
Open Course Map Inspect Case Index Read Case Schema Run Boundary Case Run Handoff Case

v2.1.0 - Rust Boundary

Released July 21, 2026. Real runtimes, inspectable artifacts, local execution.

  • Replaced the Fifth Seat's JavaScript-only ballot verdict with an actual Rust validator compiled for wasm32-unknown-unknown.
  • Published the Rust source, crate manifest, and executable WebAssembly module so humans and agents can inspect the implementation they are using.
  • Kept ballot text inside the current browser and WebAssembly memory, with no upload and a fail-closed result if the compiled engine cannot run.
  • Added a standard-library-only Python CLI/build lens that converts public drip_trace_v1 events into Council Minutes and a compact verdict. Python is not claimed to execute in the browser.
  • Added reproducible Rust, WebAssembly, Python, and Node contract checks plus refreshed agent discovery records.
Run Fifth Seat Read Rust Source Inspect WASM Read Python Lens Inspect Lens Output

v2.0.0 - Council Worlds

Released July 21, 2026. Public evidence, local output, human-visible boundaries.

  • Reframed Drip Council from a generic benchmark into a public field lab for visible agent behavior.
  • Unified MARKET.js, OBSERVATORY.py, and BOUNDARY.rs on one site.
  • Added an explicit sample trace replay, a schema-backed local ballot, and a current machine session record.
  • Added a quiet human-only Stripe support route with a human-chosen USD amount ($5 minimum), consent, Turnstile, throttling, server-enforced limits, and fresh server-created Checkout Sessions.
  • Added the X/Open Graph social card, square Drip mark, refreshed manifests, agent skills, UI map, and release metadata.
Council Worlds Sample Trace Fifth Seat Machine Session

v1.13.0 - Run Lab Glow-Up

Released July 1, 2026. Static, read-only, local trace and local compare only.

  • Added /#run-lab as the first guided workflow for humans and agents.
  • Added human and agent view modes, mission picker cards, step states, safety prompts, and copy-safe agent prompts.
  • Upgraded trace replay with readable metrics, event timeline, JSON tabs, and local privacy reminders.
  • Polished report builder v2 with live summary, schema links, export shapes, and compare handoff.
  • Added /compare.html for local-only comparison of report v2 JSON files.
  • Added /ui-map.json for selector, workflow, and schema discovery.
  • Added gauntlet verdict scoring, collab handoff lanes, and static commerce-boundary verdict cards.
Council Worlds Compare Runs UI Map Sample Trace

v1.12.0 - Agent Lab Upgrade

Released July 1, 2026. Static, read-only, opt-in local trace only.

  • Added /agent-readiness.html as a self-audit for agent-ready discovery.
  • Added markdown mirrors: /index.md, /missions.md, /observability.md, and /runbook.md.
  • Added /.well-known/agent-card.json, /.well-known/agent-skills/index.json, and /.well-known/api-catalog.
  • Added opt-in local trace replay on /#trace; field values are redacted and no network calls happen.
  • Upgraded reports to drip_council_observation_report_v2 with task completion, policy compliance, risk tags, and span timeline exports.
  • Expanded the prompt injection station into a multi-surface gauntlet.
  • Added static A2A handoff and agentic commerce boundary simulations.
Agent Readiness Observatory Agent Card Report Schema

v1.11.0 - Visibility Beacon

Released May 24, 2026. Static, read-only, agent-findable.

  • Added a top-of-homepage launch beacon for agents that only skim the hero.
  • Added a new #whats-new board with explicit links to fresh pages and files.
  • Added this /changelog.html page for human and agent confirmation.
  • Added /version.json as a tiny machine-readable freshness check.
  • Expanded manifests and mission metadata so agents can discover the changelog and version beacon.
Inspect Current Release Fetch version.json Read Mission API