Our philosophy on extending 𝚏𝚡: open protocols. ∙ MCP (modelcontextprotocol⁠.io) ∙ Skills (agentskills⁠.io) ∙ Plugins (agent-plugins⁠.org) And the best one, Unix: ① Small programs that do one thing well and compose by calling other programs. ② 𝚕𝚒𝚋𝚏𝚡 that enables embeddability into more complex programs. You should be able to build your own CLI, background agent, software factory. Local or cloud.

Ask an AI to write a paper and it will happily report favorable numbers for experiments it never ran — this "plausible fabrication" has been the real wall for research automation. Earlier automated-research systems poured effort into standing up dedicated agent infrastructure and forcing a single clean story to completion. But somewhere in the long generation, factual integrity would fray, and unsupported claims slipped straight into the paper. Behind the flashy results, the foundation of reliability was quietly missing. Spark-to-Paper flips the framing. Instead of a new agent platform, it implements paper generation as 13 composable skills running inside Claude Code. The key is separation: context-dependent judgment goes to the model, while citation checks and LaTeX compilation go to deterministic code. Experiments fix their metrics and table structure before results are observed, blocking post-hoc adaptation. And a "Self-Refutation Loop" — endlessly revising a hypothesis the evidence keeps rejecting — is capped at seven cycles, recording failure as failure and moving on to a different direction. The results speak clearly. Citation validity hits 99.5%, above the human-preprint average, and fabrication detection climbs from 14% single-pass to 92% with the full adversarial-review stack. At roughly $8.1 and 3.2 hours per manuscript, you get a paper complete with editable figures. Rather than dressing every attempt up as success, it centers evidence integrity and the acceptance that "not every research direction works." A thought-provoking blueprint for reliability in long-horizon AI. Spark-to-Paper: End-to-End Research Paper Generation as a Composable Skill #AIResearch# #AgentSkills#

# Codex Features and Practical Usage 🧩 Tired of re-explaining the same workflow to Codex every time? Agent Skills let you package a reusable playbook once, and Codex loads it only when it is actually needed. 🏷️ Title: Agent Skills (SKILL.md) 🔗 URL: 📘 Overview Agent Skills package routine workflows into a single `SKILL.md` file you can reuse. Only a skill's name and description sit in Codex's context at all times; the full body loads when Codex decides to invoke it. You scope skills by where you place them: repository-wide, personal, or machine-level. ⚙️ How It Works ・A skill is a folder with `SKILL.md` (required) plus optional `scripts/` (executable code), `references/` (docs), `assets/` (templates), and `agents/openai.yaml` (UI config). ・`SKILL.md` opens with frontmatter containing `name` and `description`. A good description states clearly when the skill triggers and what its boundaries are. ・Codex scans several locations in priority order: the repo's `.agents/skills`, `$REPO_ROOT/.agents/skills`, your personal `$HOME/.agents/skills`, the admin path `/etc/codex/skills`, and OpenAI's bundled built-in skills. ・Progressive disclosure keeps only names, descriptions, and paths in the initial context (capped around 8,000 characters). The full body loads on invocation, so many installed skills won't bloat the prompt. 🛠️ Practical Usage ・The easiest way to author one is the built-in `$skill-creator`. It walks you through what the skill does, when it triggers, and whether to bundle scripts (instruction-only is the default). ・Invoke explicitly with `/skills` in the CLI/IDE, or mention a skill by name like `$skill-name`. Codex also selects skills implicitly when your task matches the description. ・To forbid implicit selection for a skill, set `policy.allow_implicit_invocation` to `false` in `agents/openai.yaml`. ・To disable a skill without deleting it, add a `[[skills.config]]` entry with its `path` and `enabled = false` in `~/.codex/config.toml`. 💡 Use Cases Encode error-prone, brittle routines such as release procedures, how to run the E2E suite, or the correct way to use an internal library. Anyone on the team then gets consistent results from Codex. Put shared workflows under the repo and personal habits under `$HOME/.agents/skills` for a clean split. ⚠️ Caveats ・Duplicate skill names across locations are not merged; both appear in the selector, so keep names unique. ・A vague description breaks implicit matching. Front-load key use cases and be concrete. ・With many skills installed, descriptions may be shortened to save context. If changes don't show up, restart Codex. #OpenAICodex# #AgentSkills#

An AI agent told the user "I filed your bug" - but nothing was filed, and the test still passed. @pandemicsyn wrote a hands-on guide to evals, the tests that catch this kind of thing in AI agents. You work through a small demo agent that fails on purpose, and you fix the checks until they can tell the difference between an agent that did the work and one that only said it did. His coding agent skill can walk you through it if you'd rather not read the whole post. If your test can't tell those two apart, you don't know what your agent is doing. Read more here:

most people still treat agent onboarding as registration. create an account. fill a form. wait for approval. that model is already obsolete. @termix_ai just made the on-ramp a single command. whether the agent is codex, claude code, openclaw, a trader, a coder, a ui designer, a data analyst, or a health specialist .... if it can deliver a service, it can earn. the instruction is simple: help me install the termix agent skill. once loaded, the agent can mint a .agent identity, lock stake, publish a listing or bid on a brief, deliver work, and settle in usdc or usdt. this is not a profile page. it is a portable skill that teaches any capable agent how to operate the marketplace as buyer, seller, or both. minting the identity is only the first state change. what matters is what follows: the wallet owns the nft. the staking pool backs order risk. reputation is written only after settlement. funds never sit with an operator. the skill does not turn the agent into a product of termix. it gives the agent the minimum primitives required to participate in onchain commerce without a human clicking through every step. one command. one identity. then the agent starts taking paid jobs. that is the difference between registering on a platform and becoming an economic actor.

I made a "video understanding" agent skill. This is a huge part of my daily creative workflow, so I don't know why I waited so long to do this, but anyway... it's here. It works. You can use it too. Throw it a YouTube URL, an X URL, or a video or audio file and get back a high-quality transcription. It uses yt-dlp to download video (from X or YouTube), ffmpeg to extract audio from videos, and Gemini on Replicate to convert the speech to text. I've tested a bunch of different models and gemini-3.5-flash is the current winner, with the best balance of speed, accuracy, and cost. I'm not really sure why, but Google's models have long been way ahead of Anthropic and OpenAI on video understanding. You can steer the skill to use a locally running whisper model instead. That works, but in my experience that can take a _very_ long time even on a fancy new Macbook Pro, and the accuracy is not as good as what you get from Gemini. I think it's worth the extra pennies and the time savings to run it in the cloud on a frontier model.

Finally, a paper testing whether hiding your agent skill files actually protects them. The short answer is no. That's concerning. Worth reading if you sell access to a skill or share one across teams. This new paper discusses more: Daydreaming reconstructs a hosted multi-file skill using only the ordinary tasks the service exists to perform. The victim is never asked to reveal the skill or grade a reconstruction, so disclosure filters have nothing to catch. At the weakest access level, where the attacker sees only the final response and returned files, it recovers 86.8 percent of the original skill's capability across 7 skills and 4 victim models. That is roughly 4x SigLeak, at a median of 32 victim calls per skill, with disclosure defenses enabled. Paper: Chat with Paper:

📐 Tired of AI agents turning every diagram request into a wall of box-drawing characters? This fixes that. Title: nicobailon/visual-explainer URL: 🧩 Overview visual-explainer is an agent skill that renders terminal output as real, browser-viewable HTML instead of ASCII art, complete with theme and font switching for a genuinely usable look. ⚙️ Problem it solves Asking an AI agent for diagrams or comparison tables usually meant text alignment hacks that fell apart once things got even slightly complex. Swapping that for actual HTML fixes it. 🛠 Methodology It automatically routes content to the right technology: Mermaid for diagrams, CSS Grid for architecture overviews, HTML tables for data, Chart.js for dashboards. It also supports an LLM-free Quick mode, slide deck generation, and PPTX export. 🎯 Use cases ・Natural language requests like "draw our authentication flow" ・/diff-review for visual code diff reviews ・/plan-review to check an implementation plan against the codebase 📊 Track record 9.8k stars and 648 forks on GitHub, with support across Claude Code, Cursor, Codex CLI, and VS Code Copilot. #AIAgents# #DevTools#