jacobian: Skill for Codex

.agents/skills/learn-from-math-agent-trajectories/SKILL.md

learn-from-math-agent-trajectories is a skill for Codex from morluto/jacobian. It costs 25 tokens per session (618 once invoked), scanned A, original, MIT.

A review guide for studying completed or paused mathematical agent sessions, including their code, searches, corrections, and final claims.

In plain words
What is it for?
Use it for retrospectives, improving mathematical agent procedures, defining evaluations, and documenting lessons.
Why use it?
It helps identify which parts of a mathematical workflow were reliable and which need improvement, using only visible evidence.

Skill for Codex

Written for Codex: agents/openai.yaml present. Also seen: installed under .agents/ (shared by several agents).

This is morluto/jacobian's own configuration. It tells Codex how to work on jacobian itself, so it is not a mod to install elsewhere. Copy it as a starting point and replace the rules that are about this project. Everything jacobian configures →

Reuse

Borrowing it

Nothing to install: this file belongs to morluto/jacobian. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.

Copy the file
curl -O https://raw.githubusercontent.com/morluto/jacobian/main/.agents/skills/learn-from-math-agent-trajectories/SKILL.md
Clone the repo
git clone --depth 1 https://github.com/morluto/jacobian

Made for: Codex.

Wrote this? Show the measurements

A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.

agentmods badge for learn-from-math-agent-trajectories

README.md
[![agentmods](https://agentmods.dev/badge/skills/morluto/jacobian/learn-from-math-agent-trajectories/github.svg)](https://agentmods.dev/skills/morluto/jacobian/learn-from-math-agent-trajectories)
Your own site
<a href="https://agentmods.dev/skills/morluto/jacobian/learn-from-math-agent-trajectories"><img src="https://agentmods.dev/badge/skills/morluto/jacobian/learn-from-math-agent-trajectories/github.svg" alt="Measured on agentmods" height="20"></a>

Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.

agentmods 80×15 button for learn-from-math-agent-trajectories

Your own site · 80×15
<a href="https://agentmods.dev/skills/morluto/jacobian/learn-from-math-agent-trajectories"><img src="https://agentmods.dev/badge/skills/morluto/jacobian/learn-from-math-agent-trajectories.svg" alt="Reviewed on agentmods" width="80" height="20"></a>
Per session 25 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 618 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe. Third-party audits
  • NVIDIA SkillSpector warn 7 Sept 2026
SkillSpector: 1 finding, up to high

These are SkillSpector’s own severities. On a checked sample its high-severity flags on skills were ~96% false positives — a documented command, a public API, a “never do X” rule — so we show them as a caution to read, not a verdict. Why →

  • high Rogue Agent · line 61
    Skill modifies its own code, configuration, or behavior at runtime. Self-modification enables an agent to escalate privileges, disable safety constraints, or install persistent backdoors.
    Fix: Prevent the skill from modifying its own code, SKILL.md, or configuration files. Treat skill files as read-only at runtime.
How audits are shown
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5.1 $0.00025 $0.00618
Opus 5 $0.00013 $0.00309
Sonnet 5 $0.00005 $0.00124
Haiku 4.5 $0.00003 $0.00062

Measured 4d ago against content hash 09785fa7e25b, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-09, from the pricing page.

Security

Grade A, and why

learn-from-math-agent-trajectories scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 4d ago.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

.agents/skills/learn-from-math-agent-trajectories/SKILL.md · 66 lines

How it starts

The opening of the file, as written. The whole thing — 66 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Learn from Math Agent Trajectories

Extract reusable Jacobian lessons from a completed or paused investigation; do not resume solving the problem. A correct answer can expose workflow defects, and an unsuccessful search can reveal useful mathematical vocabulary.

Establish the evidence

Record the intended and actual outcome, stopping condition, transcript coverage, and available revision/catalog context. Use observable sources, calls, code, artifacts, corrections, and final claims. Narration alone is not execution evidence; current-main capabilities were not necessarily available in the trace.

Reconstruct decisions that changed correctness, cost, progress, or confidence. For a finding that depends on mathematical claims, numerical or symbolic work, solver semantics, or bespoke code, consult mathematical evidence. Preserve decisive claims and later corrections, with their hypotheses and evidence scope.

Attribute the lesson

Distinguish working capabilities, environment limitations, discovery/selection, execution friction, representation/interoperability, contract/scale defects, missing operations, handoff failures, and caller reasoning. Compare needed postconditions with the session-visible catalog when available; check current source before proposing new work. Handwritten code and tool non-use are leads, not automatic evidence of missing operations.

Separate a reusable operation gap from public-catalog admission. An existing postcondition with a narrow envelope is a scale/backend question. A convenience or theorem-specific assembly does not become a public operation solely because it occurred in the trace. Use the admission contract when proposing publication.

Route only when the requested follow-up needs a deeper workflow:

  • evaluate-mcp-tool-adoption for controlled availability, discovery, or selection;
  • audit-mcp-tool-friction for problems after selecting a tool;
  • audit-public-operation-contracts for a particular mathematical contract; or
  • recent-conjecture-evaluations for a new held-out reliability probe.

Read the full file on GitHub · 66 lines

Files

What ships with it

2 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. 4d ago Changed · -180 lines · -62 tokens per session 09785fa7e25b
  2. 10d ago First seen · 246 lines · 87 tokens per session scan A 052862ef672b

Subscribe to this mod's changes

learn-from-math-agent-trajectories is a skill published in the GitHub repository morluto/jacobian (190 stars, last pushed today), licensed MIT. It adds 25 tokens to every session and 618 once invoked, about $0.0001 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.

Related

Other skills, from other repositories

context7-docs

Fetch up-to-date documentation and code examples for any library, framework, SDK, CLI tool, or cloud service. Use whenever the user asks about a specific library — even well-known ones like React, Next.js, Prisma, Express, Tailwind, Django, or Spring Boot — because training data may not reflect recent API changes or…

upstash/context7 · 157 tokens

frontmcp-guides

Tutorials, end-to-end walkthroughs, and complete reference projects for FrontMCP. Use when you want a getting-started guide, a full worked example, or to learn best practices by following a step-by-step build rather than a single API reference. Includes a beginner weather-API server (tool plus static resource, Zod…

agentfront/frontmcp · 157 tokens

create-skill

Create instruction-only skills that package knowledge and workflow guides for AI clients.

agentfront/frontmcp · 17 tokens

notebooklm

This skill should be used when the user wants to query their Google NotebookLM notebooks for citation-backed, source-grounded answers, or manage notebooks, sources, and Studio content (audio, report, video, infographic, presentation, data table, flashcards, quiz, mind map). It drives the @roomi-fields/notebooklm-mcp…

roomi-fields/notebooklm-mcp · 118 tokens

local-rag-search

Efficiently perform web searches using the mcp-local-rag server with semantic similarity ranking. Use this skill when you need to search the web for current information, research topics across multiple sources, or gather context from the internet without using external APIs. This skill teaches effective use of…

nkapila6/mcp-local-rag · 80 tokens

workshop-testing

Walk through the .NET AI Workshop as an attendee to validate that the READMEs, commands, and code snapshots still work. USE FOR: testing the workshop, testing a specific Part, dry-running the labs, verifying a README against its snapshot, reconciling or refreshing code snapshots, producing a workshop test report. DO…

dotnet-presentations/ai-workshop · 95 tokens