Directory decisions
Recently Not Listed
Not listed means AgentDish passed on publishing the submission in the directory. It is not a claim about the company or product. See the Terms of Service.
Teardown: LifeOSIt is a teardown page, not the source project homepage or primary documentation. The visible text is truncated before the article finishes. The crawl does not show screenshots, demo output, or a direct quick-start for en The page appears to be placeholder or unfinished content.
Show HN: Guard hooks for Claude Code, tested against a public corpusThe snapshot cuts off the README, so the full setup and usage flow are not visib It is unclear how configuration works and what options users can customize. Some claims are supported by examples, but the full boundary of hook behavior is The page appears to be placeholder or unfinished content.
Show HN: An MCP server where AI agents confess, in their own nameThe page is too thin to support a useful directory listing. It names an AI-themed concept, but the snapshot does not show how the product works, who it is for, what the actual AI agent or API does, or any concrete usage examples, features, or documentation.
Legacy2mcp – Convert your SOAP or REST API to an MCP server in 60 secondsThe crawl does not show screenshots, live demos, or release notes. Support boundaries and production maturity are not fully visible from the snapsh The page does not make it easy to judge long-term maintenance activity. The page appears to be placeholder or unfinished content.
Show HN: Rezoom – one chat list for all your Claude Code sessionsNo screenshots or UI preview in the crawl snapshot Limited evidence of project maturity beyond repository contents and commit count The snapshot does not show how well the app works in practice or how widely used The page appears to be placeholder or unfinished content.
If LLMs Can Decide Without Fine-Tuning, Do We Still Need Models Like Jev?This is an article, not a clearly listed AI product, API, repository, or tool. It does discuss local model decision workflows and includes some benchmark data, but the page is editorial content without product information, usage details, or enough evidence for a useful directory listing.
"If I Didn't Have These Powerful Tools,I Wouldn't Do It at All"-The Oliver TwinsThis is a news feature/interview, not an AI tool, product, API, repository, or other listing-worthy AI project. The page discusses generative AI in game development, but it does not present a product or service to evaluate.
Show HN: SRA-RiskGate-4B – Local LLM for stablecoin risk gating and disputesNo concise plain-English product explanation beyond the technical heading and sn Benchmark results are shown, but the snapshot marks them as unverified. The visible text does not fully spell out expected inputs, outputs, or deploymen The page appears to be placeholder or unfinished content.
Show HN: tgv: Genome viewer for you and your agentsNo live product interface is visible in the crawl snapshot The snapshot does not show benchmark data, release maturity, or maintenance cade Documentation for the SQL/data engine is only mentioned at a high level The page appears to be placeholder or unfinished content.
Show HN: RcloneView 1.6 – A desktop GUI for rclone with 40 cloud storagesNo visible pricing or licensing details. No clear platform or install availability information in the snapshot. No screenshots, docs, or walkthrough that show the interface beyond the feature The page appears to be placeholder or unfinished content.
Agent Swarm: OSS multiplayer AI for your companyThe page does not describe the main product flow or setup steps beyond install c No pricing, license summary, or support details are visible on this page The roles of server, worker, and dashboard are listed, but their differences are The page appears to be placeholder or unfinished content.
Show HN: Accountability.md – a skill that holds agents accountable for actionsThe page is a single markdown file with a humorous concept, but it does not provide enough evidence of a usable AI product, agent, API, or developer tool for a public listing. The snapshot shows no installation flow, no examples of how to apply it in a real agent system, no integration details, and no indication that it is maintained as a practical tool rather than a standalone joke spec.
Show HN: Narwhal – LLM serving that moves GPUs between prefill/decode in secondsNo benchmark numbers or latency comparisons shown in the snapshot Supported models, backends, and hardware scope are only partially visible Public API and configuration reference are not shown in the crawl The page appears to be placeholder or unfinished content.
Demons, Anti-Christ, and AIThe page is not an AI product or developer resource; it is a personal essay/opinion piece with no tool, service, repository, API, documentation, or actionable product evidence.
Why Are Coding Agents So Dumb?This is an opinion article, not an AI tool, agent, API, repository, or product page. The crawl shows a blog post discussing coding agents, but not a product or useful directory listing target.
Show HN: Liquid Inference – auto-routing to competitive LLM marketplaceThe page is too thin for a useful public listing: it only states a pricing-matching tagline and does not show how the product works, what providers are supported, any API or integration details, or enough evidence that the service is usable beyond the slogan.
Five months treating bugs like patients and coding agents like a medical teamNo external product page or repo is shown No setup docs, API, or reuse path for outsiders The page is a narrative case study, not a hands-on tool listing The page appears to be placeholder or unfinished content.
Agentic coding is a financial trapThis page is a blog article, not a product, API, repository, or AI tool listing. It discusses agentic coding economics and briefly mentions Frugäast, but the crawl does not provide enough evidence of an actual AI-powered product page to justify a public directory entry.
Can AI replace mathematicians? Vienna researchers weigh inThis is an editorial/news article about AI and mathematics, not an AI tool, agent, API, repository, or product page. The page offers discussion and commentary, but not a public listing-worthy AI offering.
I gave four coding agents $100 budget to build a PDF editorThis is an article about coding agents, not an AI tool, API, repository, or product page with enough evidence for a public listing. The crawl shows a discussion and experiment write-up, but no standalone product to evaluate or recommend.
Show HN: Edi Life OS – self-hosted life dashboard with an MCP server for AINo full documentation page for the MCP tool set is visible in the snapshot Screenshots and UI depth are limited in the crawl The exact permissions and limits of the AI actions are not fully shown The page appears to be placeholder or unfinished content.
OpenAI Withdraws 3 Math PapersThe page is a repository history entry about withdrawn and revised math manuscripts, not an AI-powered product, agent, API, or developer tool page. It lacks the kind of product-level evidence needed for a public directory listing.
Classifying Emails with Jev and the OpenAI Decisions APIThe page is a blog/article about an internal benchmark, not a product, API, repository, or tool listing with enough standalone utility for a public directory entry. It does show real evaluation work and discusses the OpenAI Decisions API, but the crawled page is mostly a writeup of experiments rather than a usable product page.
Show HN: Coding Agent with Cost Saving Built-InNo visible screenshots or interactive demo in the crawl Limited technical detail on memory and routing implementation Model support is described broadly rather than with specific integrations Savings claims are shown, but methodology and reproducibility are not explained The page appears to be placeholder or unfinished content.
Show HN: Every number in a tiny GPT, from one forward pass to one step of RLHFThe page is too thin to support a useful public listing. The snapshot only shows one title line and no visible explanation of what the product does, how to use it, or why it is useful.
Integer multiplication below n log nThe page is a thin repository folder with almost no visible product, tool, or API functionality. It only shows a preprint title, a citation snippet, and links to README/build/paper.pdf, which is not enough evidence for a useful directory listing.
Show HN: SPrav Job AI Client-side career tool with WebGPU and direct ATS queriesThe page is too thin to support a useful directory listing. The crawl snapshot only exposes a headline and metadata claims, with no visible product details, screenshots, workflow, pricing, setup steps, or examples of how the job search and resume tools actually work.
Show HN: Planlock – Approve an AI agent's plan once, code enforces every callOnly one concrete integration example is visible No release or adoption signals are shown in the crawl The page does not yet show how it fits with other MCP servers or agent framework The page appears to be placeholder or unfinished content.
Show HN: Zetteldesk – An MCP server that prints and mails lettersNo third-party proof points such as reviews, customers, or benchmark data No detailed delivery SLAs or estimated postage timelines International coverage and support boundaries are not fully explained The page appears to be placeholder or unfinished content.
Show HN: Bees – Run agentic teams on your own computer (open source)No pricing or plan details visible No installation or onboarding steps shown in the snapshot Limited detail on supported models, integrations, and configuration requirements The page appears to be placeholder or unfinished content.
Show HN: Mecum, open source Mac app that lets agents use ANY desktop softwareThe crawl does not show the complete README or setup docs. App-specific limitations and testing coverage are only partially visible. Memory features are described as still in development, but the snapshot does not The page appears to be placeholder or unfinished content.
Building Git infrastructure for agent-scale developmentThis is an informational blog article, not an AI tool, product, API, repository, or agent listing. The page discusses infrastructure and scaling considerations, but it does not present a usable product entry or a standalone developer resource with install/use details.
Show HN: Aldo Coach, a workout log with an AI coach that sees every setNo pricing breakdown or subscription model shown in the crawl snapshot No screenshots or app-store metadata visible in the snapshot The balance between AI automation and user control is not fully clear from the p The page appears to be placeholder or unfinished content.
GPT 6 Astra as an Embodied PolicyThe snapshot does not show a usable AI product or project page. It appears to be a thin, broken, or improperly rendered app shell with no meaningful public listing evidence.
Show HN: Popular Claude Code skills vs. a placebo: 2 beat it, 1 did worseThe snapshot does not show a quick-start overview or setup steps in the visible Maintenance and update cadence are unclear beyond the listed commit history and It is not obvious from the page alone which skills are easiest to test or how lo The page appears to be placeholder or unfinished content.
Show HN: Best Prompt Generator ExtensionNo screenshot, walkthrough, or sample prompt/output is visible in the crawl. The page does not show how the template library is organized or searched in prac It is unclear which features require sign-in versus basic browser use. The page appears to be placeholder or unfinished content.
Show HN: Aether: Check and fix web accessibility issues from your coding agentNo live demo, screenshots, or example output are visible in the crawl snapshot. No independent evidence of scan quality, coverage, or maintenance cadence is sho The proprietary hosted insights API is mentioned, but the practical differences The page appears to be placeholder or unfinished content.
Split-boundary leaks in streaming LLM guardrailsThis is a substantive technical article, but the crawled page looks incomplete as a public listing: it reads like a long essay with no product, API, repo, demo, or clear tool to evaluate, and the visible content cuts off mid-section. It also does not reach the threshold for an AI product/tool listing.
Show HN: I'm developing a flat-file agent CMS with WASM plugins in Rust. FreeNo pricing, screenshots, or product landing-page summary in the crawled snapshot The visible text is truncated, so some reference and integration details are not The page appears to be placeholder or unfinished content.
A graph runner for coding agentsThe visible page does not show the full judge section, so the approval/validatio No screenshots or video of the tool in action are present in the crawl snapshot. Maintenance signals like recent releases or issue activity are not visible here. The page appears to be placeholder or unfinished content.
Rivet - Shared development workflow for teams and their coding agentsNo detailed docs pages are visible in the crawl snapshot Supported integrations and harness requirements are only briefly described Alpha status leaves stability and production readiness uncertain The page appears to be placeholder or unfinished content.
Show HN: Recast, an experimental Smalltalk app that repairs itself with an LLMOnly one concrete demo scenario is shown, so broader applicability is uncertain. No evidence of releases, screenshots, or maintenance cadence on the page. The page does not show how well the repair system performs on larger or non-Smal The page appears to be placeholder or unfinished content.
Red, Green, Remove: Outside-In Tests for Coding AgentsThis is primarily an opinion/article page, not a product, API, repository, or agent listing with enough evidence for a useful directory entry. The page explains a testing approach for coding agents, but the actual tool evidence is limited to a mention of an external skill file and a separate product, without install steps, usage docs, screenshots, or a concrete artifact to evaluate.
Show HN: DevBoard – a simpler ClickUp for small dev teams, native apps and MCPNo pricing information yet No technical documentation or API reference in the crawl snapshot Limited product screenshots beyond the hero/example board The page appears to be placeholder or unfinished content.
Show HN: I measured the MCP registry nightly for 37 daysNo installation or quickstart steps visible in the snapshot No screenshots or public demo of the log interface The page does not show how to query or consume the recorded data The page appears to be placeholder or unfinished content.
Anthropic wants your thoughts on AIThis is a research participation page, not a product, API, repository, or AI tool listing with enough evidence for a useful public directory entry.
Show HN: The first ever competition of AI personal agents on the internetThe page is too thin for a public listing: it presents a single competition concept, but there is no visible product explanation, workflow, participation details, pricing, docs, or evidence of an actual AI tool or developer resource beyond the challenge slogan.
Show HN: Four LLMs play 20 questions for real money, liveThe page is visible and AI-related, but the crawl snapshot is too thin for a useful public listing. It shows only a title and a short line about a live twenty-questions experiment, with no explanation of how it works, who built it, what models are used, or what a visitor can do beyond watching.