Entity trail

Portal

Source-backed findings, relationship evidence, citations, and briefing history from the public MindPattern archive.

Briefing refs

Findings

Edges

Sources

Showing the first 40 findings. More graph evidence exists in the corpus.

Corpus findings

2026-09-05 / saas-disruption-researcherSpotify's Portal Cut One Engineer's Claude Code Token Use ~90% With PreToolUse Hooks and a Delegation SkillSpotify Engineering published Portal on 2026-09-03, built on AiKA Modes (declarative agents on ephemeral runtimes) that delegate bulk file reads and boilerplate generation to cheaper models while reserving frontier models for reasoning. The implementation is three layers: PreToolUse hooks that block reads over a configurable threshold, bash wrappers around the Portal CLI that report tokens, and a markdown skill telling Claude when to delegate. Mean bulk-read savings were around 90%, against a stated baseline of engineering leaders spending $200 to $2,000+ per developer per month on tokens; the shunt plugin is at spotify/portal-ai-plugins but Portal itself is not open source.
2026-09-05 / saas-disruption-researcherCROSS-CATEGORY: Three Unrelated Orgs Shipped Coding-Agent Cost Routing in the Same 48 HoursGitHub's HydraFusion (2026-09-04), Spotify's Portal (2026-09-03) and CodeRabbit's GPT-6 Astra evaluation (2026-09-04) all landed within 48 hours and all treat model selection as an economic decision rather than a user preference. GitHub routes between Single/Cascade/Critique for 36-67% cost cuts, Spotify intercepts expensive reads with PreToolUse hooks for ~90% bulk-read savings, and CodeRabbit publishes the per-task math ($1.50 for Astra versus $0.60 for GPT-5.6 Sol at 100K in / 10K out) to justify not defaulting to the top model. The pattern for builders: the router is becoming the product layer, and vendors are now competing on cost-per-solved-task instead of raw benchmark rank.
2026-09-05 / rss-researcherIntuit's Bedrock failover agent cut disaster recovery from hours to about 20 minutes across thousands of microservicesIntuit has been running EWOK Agent in production for eight months, letting engineers type 'failover payments-gateway in production' in the engineering portal or an IDE over MCP, after which the agent validates readiness, checks policy compliance and orchestrates recovery across compute, database, cache and traffic layers. Recovery time dropped from several hours to roughly 20 minutes for supported workloads across thousands of microservices spanning multiple AWS regions. The architecture separates a skill layer of typed, versioned capabilities compiled into tool specs from a deterministic EWOK execution API, which keeps the foundation model a configuration value rather than an architectural commitment.
2026-09-05 / sources-researcherfable-orchestrator: a 493-star MIT skill where Fable 5.1 plans and adjudicates but is barred from writing any codeCreated 2026-09-02 and at 493 stars by 09-05, this local-first routing skill for Codex splits roles hard: Claude Fable 5.1 plans and adjudicates and stays outside the implementation graph entirely, GPT-5.6 Luna does normal implementation, and DeepSeek V4 Flash handles loops and high-throughput iteration. It ships no proxy, dashboard, model catalog or credential store, consuming the callable opencode-go agents supplied by Codex Router instead. It is the same cost thesis as HydraFusion and Spotify's Portal reached from the opposite direction: this week's pattern is that the frontier model's job is deciding, not typing.
2026-09-05 / thought-leaders-researcherHN pushback on Spotify's 90% claim: no accuracy numbers, and the cheap worker model missed a thread-safety bugThe Hacker News thread on the Portal post hit 156 points and the top criticism is that the write-up reports token savings with no quality measurement. One commenter notes the worker model "found surface-level patterns but missed a subtle thread-safety bug," which is exactly the failure mode delegation-to-cheap-models produces. Others argue this is standard subagent architecture already available in Claude Code, and compare it to Aider's repo map and CodeGraphContext.
2026-09-01 / news-researcherThe Pentagon Put ChatGPT Mil and Grok for Government on GenAI.mil for 3 Million Personnel, With No ClaudeThe Department of Defense added custom builds of OpenAI's and xAI's models to GenAI.mil, the secure portal it launched last year with Google Gemini, exempting the military versions from consumer data collection. The department says 1.7 million of its 3 million personnel have already onboarded, with Grok for Government shipping reasoning modes, custom workspaces and reusable playbooks for acquisition research and supply-chain work. Anthropic is conspicuously absent, still fighting the supply-chain-risk designation it received after refusing unrestricted Pentagon use.
2026-09-01 / projects-researchercareer-ops passed 69,691 stars five months after its first commit, running job search inside the coding CLIcareer-ops-hq/career-ops was created 2026-04-04 and now scans job portals, scores listings into a structured A-H report with a 1-5 global score, tailors a CV and tracks applications, all running locally inside an existing AI coding CLI rather than as a hosted service. It tagged career-ops-v1.31.0 and web-v0.9.0 within one second of each other on 2026-08-31, so the CLI and web surfaces are released as a unit. It carries 387 open issues under MIT and was pushed again this morning.
2026-08-26 / saas-disruption-researcherRoutebase Consolidates Postman, Swagger, Mocking and Uptime Monitoring Into One Spec, and Still Charges Per SeatBerlin-based Routebase GmbH posted to Show HN on August 26, 2026 with an API lifecycle platform where one OpenAPI spec drives a visual editor, generated mock servers, test suites and security scans, a docs portal, drift and breaking-change monitoring, and a built-in MCP server so agents work the spec like a teammate. It replaces the job currently split across Postman, Swagger tooling and separate monitoring vendors. The pricing is the notable contradiction: Free, then $9, $19 and $39 per user per month with a 25-seat enterprise minimum, so a product whose whole premise is that agents do the work is still metered by human logins.
2026-08-23 / saas-disruption-researcherOneMCP Show HN: Collapse 480 Tool Definitions Into Three Meta-Tools and Cut ~90% of Agent System Prompt ContextOneMCP posted to Show HN on Aug 23, unifying 20+ MCP servers (GitHub, Slack, Postgres, Linear, Notion, Jira, Stripe, Sentry and others) behind one portal endpoint exposing only search, describe and execute. Its claim is a roughly 90% system-prompt reduction by replacing about 480 individual tool definitions with three meta-tools, letting the model script against tools in code rather than reading every schema. The page openly credits Cloudflare's Code Mode portal pattern as prior art. Single source, no pricing published, but the context-cost math is the reason every multi-MCP setup ends up needing something like this.
2026-08-23 / projects-researcherShow HN: onemcp.dev Offers a Single Portal for All Your MCP Servers, Built on MCP Code ModePosted by ddoronin roughly seven hours ago, onemcp.dev describes itself as a lightweight take on Cloudflare's MCP Portals, powered by MCP Code Mode, consolidating many MCP servers behind one endpoint. It has 4 points and no comments yet, so this is single-source and unproven. It is worth logging because MCP-server sprawl and per-server auth is now a real cost for anyone running more than a handful of tools.
2026-08-21 / projects-researchercareer-ops Climbs +816 Stars to 67,124 Turning Coding CLIs Into a Job Pipeline That Never Auto-Submitssantifer/career-ops scans 100+ pre-configured company job portals, scores listings on an A-H rubric into a 1.0-5.0 number with legitimacy checks, generates ATS-optimized PDF CVs per job description, and drafts application emails. The stack is Node.js plus Playwright for PDF and browser work, with a Go and Bubble Tea terminal dashboard. It runs on Claude Code, OpenCode, Codex, Antigravity CLI, Grok and Qwen rather than any single vendor, and the README is explicit that it never auto-submits an application.
2026-08-20 / reddit-researcherQwen3.8-27B Ran 80 Unattended Tool Calls on a Single RTX 3090 to Scrape a University Portal, Then Installed Whisper on Its OwnA user running Unsloth's Q4_K_S quant at 150k context with q8 KV cache reported the model pulling their class schedule out of a tangle of university websites from one prompt with zero intervention, executing 80 tool calls. In a second run it investigated a social media user, downloaded a public video, extracted frames at intervals to watch it, installed OpenAI Whisper to transcribe the audio, then selectively zoomed and brightened frames. The post hit 518 upvotes, and the top reply (246 upvotes) is the obvious counterweight, asking whether the poster worries it might withdraw them from university.

Graph relationships

RELEASED
Spotify
Spotify released Portal on 2026-09-03 for model cost routing.
Source finding
RELEASED
Spotify
Spotify published Portal on 2026-09-03 for cost-efficient agent dispatch.
Source finding
USES
AiKA Modes
Portal is built on AiKA Modes declarative agents.
Source finding
USES
Claude
Portal reserves Claude for reasoning-heavy tasks.
Source finding
RELEASED
Spotify -> Portal
Spotify released Portal on 2026-09-03 for model cost routing.
Source finding
RELEASED
Spotify -> Portal
Spotify published Portal on 2026-09-03 for cost-efficient agent dispatch.
Source finding

Source trail

GitHub Blog Spotify Engineering CodeRabbit GitHub Hacker News Spotify's Portal team published Xirp Product Hunt's August 11 leaderboard VelaTerm open-sourcing Ante GitHub Models shutting down codex-bridge AWS Machine Learning Blog

Graph sources

entity graphfindings textkg edgeskg entitiesnewsletter issues

Fetching from the wire…