Entity trail

Claude Skills

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

2026-08-22 / skill-finderClaude Code 2.1.239 adds /claude-api upgrade for the anthropic 0.x to 1.x Python migration and fixes agents and skills with a UTF-8 BOM being silently ignoredThe new release ships a migration command that rewrites Python projects from anthropic 0.x to 1.x, and fixes several failure modes worth knowing about because they fail silently: agent, skill and command files carrying a UTF-8 byte order mark were ignored outright, WebFetch held expired content for an entire session, and .worktreeinclude patterns beginning with **/ matched nothing. ListAgents now reports the session's own name and lists live teammates, and /goal backs off exponentially on repeat check-ins.
2026-08-22 / vibe-coding-researcherPattern: Agent Runtimes Are Consolidating Under Neutral Foundationsgoose's move to the Agentic AI Foundation at the Linux Foundation follows the Agent Plugins 1.0 spec published by a vendor-neutral technical steering committee on 2026-08-06 and Anthropic's release of the Agent Skills format as an open standard. The practical effect for builders is that skill and plugin directories are starting to be portable across Claude Code, Cursor, Codex and others rather than locked to one client. The caveat that keeps surfacing in practitioner writeups is that portability is not efficacy: focused skills with at most three modules reportedly outperform exhaustive bundles.
2026-08-22 / vibe-coding-researcherTip: BOM-Prefixed Markdown Silently Disabled Your Agents, Skills and CommandsClaude Code 2.1.239 fixes agents, skills and slash commands whose `.md` file begins with a UTF-8 byte order mark being silently ignored, with no warning that the file existed. If you have authored `.claude/` markdown on Windows or with an editor that writes a BOM, files you believed were active may never have loaded. Same release also fixes `.worktreeinclude` patterns starting with `**/` matching nothing when the target lived in a gitignored directory.
2026-08-21 / reddit-researcherAnthropic Took Computer Use, the Skills API and the Files API to GA, and Added a Browser Use Tool That Reads Page StructureAnthropic announced general availability on August 20. Computer use now takes several actions per turn (click, type, key, screenshot) rather than one per round trip, which early-access customers measured as 20-40% fewer round trips per task. The new browser use tool gives Claude the page structure instead of pixels so automations survive layout changes, the Skills API adds versioning with per-request pinning to a version or `latest`, and the Files API adds expiring files, 5x higher rate limits at 500 RPM, and 1 TB of storage per org.
2026-08-21 / vibe-coding-researcherAgent Skills and the Skills API Go GA, No Beta Header RequiredAs of Aug 19, Agent Skills and the `/v1/skills` endpoints are generally available on the Claude API. Messages API requests that load Skills through the `container` parameter no longer need the `skills-2025-10-02` beta header, though requests still sending it keep working unchanged. The versioned-instruction-package format that started as a Claude Code convention is now a first-class API primitive you can upload, version and mount per session.
2026-08-20 / skill-finderClaude Code adds a built-in Concise output style and cuts the claude-api skill from 200k tokens to 25kVersion 2.1.237 (Aug 20) ships a Concise output style, selectable under Output style in /config, that makes Claude lead with results and skip preamble and narration while doing the same work. Version 2.1.234 separately dropped the context cost of the bundled claude-api skill from over 200k tokens to about 25k by loading reference docs on demand. Both are worth copying in your own skills, load references lazily rather than inlining them, and put verbosity in a switchable style instead of repeating it in every prompt.
2026-08-20 / github-pulse-researcherbook-to-skill Turns a Technical Book Into an Agent Skill and Claims 24x to 51x Fewer Tokens Than Dumping It Into ContextLeutenegger/book-to-skill (1,209 stars in 6 days, roughly 202 a day) distills a PDF, EPUB, DOCX, MD, HTML, RTF or MOBI file, folder or glob into a single skill usable from GitHub Copilot CLI, Amp or Claude Code, targeting the open Agent Skills standard rather than one vendor. The headline claim, measured on real books, is 24x to 51x fewer tokens to answer one question than putting the book in context. Install is 'pip install -e .' then 'book-to-skill install'.
2026-08-20 / rss-researcherAnthropic moves Files API, Agent Skills, and Admin API to GA and adds domain allowlists for Managed AgentsOn August 19 the Claude Developer Platform promoted three betas to general availability: Admin API user-management endpoints for Enterprise orgs, the Files API with expiration, pagination, and 1 TB of storage per organization, and Agent Skills with the /v1/skills endpoint. Managed Agents gained allowed_domains and blocked_domains controls over web access, self-hosted sandbox support for memory stores, and a rebuilt Console session viewer with a timeline minimap and an Inspector panel. The day before, the Workbench was renamed Playground with full Messages API support and runnable templates.
2026-08-20 / thought-leaders-researcherAdam Rackis walks through Matt Pocock's 'Skills for Real Engineers', the 224K-star skill suite built around interrogating your own promptRackis published a hands-on writeup on August 18 (promoted on X August 19) of Pocock's skill pack, installable via `claude plugins install mattpocock-skills`. The centerpiece is `/grill-me`, which formalizes the "did I miss anything?" prompt into a structured interrogation before any code is written, paired with `/to-spec` to freeze a conversation into an issue, `/to-tickets` to split it into several, `/teach` to generate a lesson, and `/setup-matt-pocock-skills` to configure where issues and ADRs land. The repo is MIT-licensed at 224,848 stars and 19,319 forks, last pushed August 20.
2026-08-20 / thought-leaders-researcherAnthropic ships 'discernment-nudge', a skill whose only job is deciding when Claude should tell you to double-check itDaniel San flagged the skill on August 20, calling it the "best trigger logic writing I've read this year" and noting the section on when NOT to fire is about three times longer than when to fire. The SKILL.md confirms it: the skill appends 2-3 follow-up questions after substantive answers users will act on (estimates, consequential advice, factual claims), at most once per conversation, and carves out trivial lookups, educational explanations, format conversions, code the user will run, creative writing and any request where the user already asked for a review. Install is `/plugin marketplace add anthropics/skills` then `/plugin install discernment-nudge@anthropic-agent-skills`; the repo sits at 170,610 stars.
2026-08-19 / skill-finderClaude Code 2.1.233-2.1.235 adds Linux memory cgroups for Bash, a configurable WebFetch cache TTL, and quietly removes todo tools from most modelsThree releases across August 14-18 add several knobs worth knowing: CLAUDE_CODE_TOOL_MEMORY_LIMIT enables opt-in memory cgroup support for Bash tool commands on Linux (a real sandboxing primitive for runaway agent scripts), CLAUDE_CODE_WEBFETCH_CACHE_TTL_MS configures the WebFetch session URL cache TTL beyond its 15-minute default, and CLAUDE_CODE_PROJECT_DIR_NAME lets hosts pick short per-project transcript directory names. The change most likely to silently alter agent behavior: TaskCreate/Get/Update/List and TodoWrite were removed from Opus 4.8, Sonnet 5, Fable 5, and Mythos 5+ in 2.1.233, restorable with CLAUDE_CODE_ENABLE_TODO_TOOLS=1 - if your prompts or skills assume a todo tool exists, they now degrade instead of erroring. Also new: sessions continue automatically when a claude.ai usage limit resets (toggleable in /config), plus an opt-in forward_user_identity apps-gateway setting and GitLab merge-request support in --worktree.
2026-08-17 / github-pulse-researcherClaude Codex Bridge v8.6.9 Adds DeepSeek Harness as a Separate Provider Key and Refuses to Let a Pane Declare a Turn CompleteSeemSeam/claude_codex_bridge shipped v8.6.9 on 2026-08-17 (3,425 stars, two releases in one day) adding official DeepSeek Harness under a distinct Developer Preview provider key `dsh`, leaving the existing `deepseek` key pointed at Deep Code CLI. The notable engineering: DSH completion requires an exact durable `source.rpcId`, a committed non-empty assistant reply from the same turn, *and* a native turn/end(completed) event — the POSIX pane is demoted to a lifecycle/log carrier with no authority to declare prompt, reply, quiet-time or successful completion, and restore is observer-only so it never reposts a prompt. Credentials, skills, memory and caches stay inside an agent-private DSH_HOME with one-directional allowlisted copying. This is the clearest published statement yet that terminal-scraping multi-agent bridges cannot trust pane text as a completion signal.

Graph relationships

RELEASED
Anthropic
Anthropic released Claude Skills for progressive disclosure pattern.
Source finding
SUPPORTS
Claude Code
claude-skills collection supports Claude Code with 192+ skills.
Source finding
RELEASED
Anthropic -> Claude Skills
Anthropic released Claude Skills for progressive disclosure pattern.
Source finding

Source trail

talk and engineering post released Agent Skills as an open standard VoltAgent's awesome-agent-skills HuggingFace published SKILL.md files Channels viral 4-minute video Northeastern researchers Constellation Research's RSAC analysis practical guide The Hacker News OpenAI Blog Help Net Security

Graph sources

entity graphfindings textkg edgeskg entitiesnewsletter issues

Fetching from the wire…