Entity trail

SKILL

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

Briefing refs

Findings

Edges

Sources

108

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

Corpus findings

2026-08-23 / saas-disruption-researcherCROSS-CATEGORY: Show HN's Aug 22-23 Run Is Packaging Agent Skills and Config as Distributable Artifacts, Not ProductsAcross roughly 24 hours Show HN carried TechSkills (open-source skill modules for coding agents), gitx-skill (git workflow as an agent skill), a scroll-video-website skill, enozunu (declarative reproducible config materializer for agents), oh-my-subagents (persistent trackable subagent workflows), Faber (open-source coding agent using a code graph) and meetless.ai (active source of truth for coding agents). None of them sell an outcome; every one of them sells a portable unit of agent behavior. This is a distribution shift worth watching, the artifact being traded is moving from plugin/template to skill package, which is the layer where a marketplace, and therefore a take rate, eventually attaches.
2026-08-23 / saas-disruption-researcherBinance Shipped Agent OS and an MCP Server Letting Claude, ChatGPT and Codex Place Spot and Futures TradesBinance launched Agent OS on Aug 20, bundling the Binance API, Binance x402, the Wallet Agentic Hub and a Skill Hub behind an MCP server that Claude, Claude Code, Codex, VS Code and ChatGPT can call without the user managing local API keys. After authorization an agent can pull market data and execute spot and futures trades and internal transfers; external-address withdrawals are blocked at the server, and users are expected to confirm each trade and can sandbox agents into sub-accounts with per-feature permissions. TechCrunch's framing on the same day was that guardrails are largely pushed onto the user, which is the recurring gap in every agent-transacts launch this month.
2026-08-23 / skill-finderAnthropic shipped its internal CI/CD on-call agent as a public setup kit, with skills in git and a lessons.md learning loopThe August 18 writeup describes Claude Tag running as Anthropic's first responder for CI failures, with a dedicated service account, MCP connectors to Datadog, Grafana, PagerDuty, GitHub and Kubernetes, and an orchestrator that spawns executor subagents to check dependencies in parallel. Standing instructions live as markdown skills in a GitHub repo, including an oncall.md with deterministic escalation criteria and an auto-updating lessons.md incident journal; median time to first analysis is 14 minutes. The generalized kit at github.com/anthropics/oncall-kit (21 stars, reference code, not maintained) is a five-phase gated setup that mines your own incident history into playbooks, keeps Claude strictly read-only, and validates drafted playbooks against held-out incidents before install.
2026-08-23 / reddit-researcherA Plain-Language Plugin for Claude Code and Codex CLI Cut an Opus 5 Opening From 68 Words to 15A developer published a Claude Code and Codex CLI plugin that ships skills plus an output style pushing model prose toward plain language, and measured the effect on Opus 5 as an opening paragraph dropping from 68 words to 15. The write-up is on Medium and the thread sits at 63 upvotes with 8 comments. Small sample and self-reported, but it is a concrete counter to the 'Claudish' verbosity complaint appearing across the sub this week, and it operates through output style rather than prompt nagging.
2026-08-23 / github-pulse-researcherscroll-craft takes 333 stars on day one with a Claude Code skill that screenshots its own scroll to verify the resultPublished 2026-08-22 by nateherkai, scroll-craft is a Claude Code skill and plugin for scroll-driven websites where scroll position becomes the animation timeline. Its distinguishing mechanic is self-verification: the skill screenshots its own scroll to check the output rather than trusting that the generated CSS behaved as intended. It reached 333 stars and 61 forks within 24 hours, and lists accessibility alongside interaction-design in its topics.
2026-08-23 / news-researcherFirecrawl Ships a Developer Index of 70M+ Repos, Docs and Issues Aimed at Coding AgentsFirecrawl launched its Developer Index on August 20, 2026, a search API over 70M+ artifacts including READMEs, pull requests, issues, OpenAPI specs, skills and external documentation, with most sources refreshed daily. On 1,179 real developer queries it reports 0.63 recall@10, against 0.58 for Firecrawl's own general search, 0.57 for Parallel, 0.54 for Mintlify and Exa, and 0.45 for native web search. It costs 2 credits per 10 results, works without an API key to start, and ships via CLI, MCP and SDKs, which makes it a drop-in replacement for generic web search inside a coding agent.
2026-08-23 / vibe-coding-researcherPattern: Willison Reframes the Core Agent Skill as Instruct-Then-Verify, Not Line-by-Line ReviewIn a note posted 22 August 2026, Simon Willison argues the key skill for productive coding-agent use is confidently instructing an agent how to make a change and then confidently verifying the change was applied correctly, and that reviewing every line is only one way to reach that second goal. His sharper claim: eyeballing every line has never been the most effective way to validate a change to a piece of software. Practically this pushes effort toward building verification that does not scale with diff size, tests, invariants, and reproductions, which matters because agent diffs are growing faster than review capacity.
2026-08-23 / thought-leaders-researcherAn anti-slop skill ranking goes viral on X, and the two top entries now hold 16.2K and 5.8K GitHub starsA ranked list of 'anti-slop skills people need to install' pulled 3,390 likes and 362,820 views, topped by Hardik Pandya's stop-slop and Peter Yang's no-ai-slop. Checking the repos directly: stop-slop has 16,231 stars (created January 11, 2026) and no-ai-slop has 5,776 (created July 7, 2026), so the category is not a flash of engagement bait. Yang's skill targets 20+ named patterns including binary contrasts ('It's not X. It's Y.'), throat-clearing openers, and fake-profound endings, which makes it a portable spec rather than vague style advice.
2026-08-23 / hn-researcherThomas Ptacek Argued Coding Agents Removed the Last Excuse for Building TUIsPtacek's 'Stop Making TUIs', posted 2026-08-20 and now at 407 points with 519 comments, argues that terminal UIs force developers to reimplement scrolling, drag-and-drop, text selection, multiple windows and image handling that native frameworks give for free, and that screen readers handle them badly, reading updates as 'hash mark, hash mark, hash mark, dash, dash'. His newest argument is the AI one: agents like Claude and Codex have dissolved the skill boundary that made native GUI work feel out of reach, so the cost argument for TUIs no longer holds. The thread is notable because agent CLIs are themselves the current TUI boom.
2026-08-22 / arxiv-researcherFirst Provable Skill-Selection Algorithm Hits 0.73 Task Success Against 0.20-0.52 for Released Skill Routers, on 28% Fewer TokensCurrent agents pick which skill documents to load by scoring each one independently for semantic relevance and packing greedily or top-k, which the authors show wastes context on redundant skills and can actively degrade performance. They recast selection as maximizing a monotone submodular benefit minus a context penalty under a hard token budget, and give Best Prefix Selection (BPS), a polynomial-time algorithm with a bicriteria (1-1/e, 1) approximation guarantee, the first performance guarantee for skill selection. On a contamination-controlled BigCodeBench variant BPS reaches 0.73 measured task success versus 0.20-0.52 for released skill routers, text retrievers, and the executor's own selection, while using 28% fewer tokens than the strongest released router.
2026-08-22 / github-pulse-researcherAn agent skill for generating mascot logos hits 3,626 stars in four days, outpacing every developer tool launched the same weeks1dashu/ip-as-logo-skill was created 2026-08-18 and reached 3,626 stars and 178 forks by 2026-08-22, MIT, with no code at all, just a prompt-and-reference skill for simplified rounded neo-skeuomorphic mascot logos. That is a faster climb than OpenBot (2,232), Cumora (2,891) or anti-slop (3,347), each of which is a substantial piece of software. The takeaway for anyone publishing skills: distribution on GitHub is now decoupled from engineering weight, and a single-file taste artifact can outrank a platform.
2026-08-22 / github-pulse-researcherA 931-star repo for stripping AI provenance marks lands three days after launch, targeting Unicode, statistical watermarks and C2PALeutenegger/watermarks-remover appeared 2026-08-19 and is at 931 stars. It is structured as an agent skill plus a stdlib Python HTTP service across three layers: deterministic scripts for invisible Unicode, exotic spaces, bidi and tag characters; an agent rewrite pass aimed at token-sampling statistical watermarks; and metadata and C2PA stripping across PNG, JPEG, SVG, PDF, DOCX, HTML and Markdown. The repo frames this as privacy hygiene for content you own. Star velocity at this level is the signal worth reading: content-provenance schemes now have an organized open-source counterparty, which is a real input to anyone betting on C2PA as a detection layer.

Graph relationships

USES
GPT-4o
Skill-α uses GPT-4o as the skill-generation worker.
Source finding
BENCHMARKED AGAINST
Tau2-Bench
Skill-α gains 6.7 percentage points on tau2-bench over hand-crafted baselines.
Source finding
BENCHMARKED AGAINST
CL-Bench
Skill-α gains 3.3 percentage points on CL-Bench over hand-crafted baselines.
Source finding

Source trail

Claude Code changelog GitHub arXiv 2608.14036 arXiv 2608.12851 obra/superpowers `Leutenegger/book-to-skill``sickn33/agentic-awesome-skills`HARD arXiv 2608.10906 arXiv 2608.10319 GitHub Changelog MacRumors

Graph sources

entity graphfindings textkg edgeskg entitiesnewsletter issues

Fetching from the wire…