Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx agentmods add commands/afar1/fieldtheory-cli/fieldtheorygit clone --depth 1 https://github.com/afar1/fieldtheory-cliWhat 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.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5 | $0.00088 | $0.04024 |
| Opus 5 | $0.00044 | $0.02012 |
| Sonnet 5 | $0.00018 | $0.00805 |
| Haiku 4.5 | $0.00009 | $0.00402 |
Grade A, and why
fieldtheory 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 3d 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.
How it starts
The opening of the file, as written. The whole thing — 285 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Field Theory CLI
The Field Theory CLI (ft) is a self-custody bookmark tool that does two related things:
- Local X/Twitter bookmark archive — sync, full-text search, classify, visualize.
- Possibility runs (
ft possible) — take a group of bookmarks, apply them to a set of repos, and score candidate directions onto a 2x2 grid. Each scored idea is a "node" (also called a "dot") with a paragraph summary, a copiable goal prompt, and per-axis justifications. Seeds, runs, nodes, and batches are all saved as interconnected markdown files.
The feature was previously named
ft ideasand the old name still works as an alias.ft possibleis the primary vocabulary — use it in new walkthroughs, scripts, and explanations.
Everything runs locally. Bookmark storage lives at ~/.ft-bookmarks/; ideas data lives at ~/.fieldtheory/ideas/. Nothing leaves the machine.
Mental model — the possibility flow
A seed is not raw text. A seed is:
- A bookmark, or a group of bookmarks (the external context/ideas)
- + a repo or a set of repos (the codebases to apply them to)
- + a frame (the 2x2 axes that shape the debate — e.g. Leverage × Specificity)
Those three things together shape a run. A run invokes an LLM pipeline that reads the seed, scans each repo, generates candidate directions, critiques them, and scores the survivors onto the 2x2. Each scored candidate becomes a node/dot on the grid with:
- title + paragraph summary
- rationale ("why adjacent")
- axis A score + justification, axis B score + justification
- effort estimate (hours / days / weeks)
- a copiable goal prompt that can be pasted into any AI coding agent to actually build the thing
When a run spans multiple repos, the seed brief is computed once and reused across every repo (via the seed-brief cache), then each repo gets its own survey → generate → critique → score pass and its own consideration. A top-level batch_summary artifact + markdown file links them all together.
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.
- 3d ago First seen · 285 lines · 88 tokens per session scan A 4c925c1c1e26
fieldtheory is a command published in the GitHub repository afar1/fieldtheory-cli (2,018 stars, last pushed 14d ago), licensed MIT. It adds 88 tokens to every session and 4,024 once invoked, about $0.0004 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.
Other commands, from other repositories
graph
Extract prerequisite concept graph from converted course materials and write course-index/concept-graph.md.
reindex
Idempotent reindex of an existing course — rewrite retired coverage.md tier markers to canonical vocabulary and materialize errors/log.md nature/phase into the on-disk schema (header keys == data keys), in place and atomically, WITHOUT running analyze.
alt
Import an Exam Radar (OPTIMETA Alt plugin) export and fold its lecture-emphasis exam signal into the course index — radar.md, a lecture-emphasis column on coverage.md, and a gold-zone weakmap.
blind
Strategy-level blind drill on a known HW or example problem. User describes approach in prose (no math typing); Claude verifies against solution then saves clean reference to derivations/.
doctor
Diagnose the paideia install + course workspace (Python, poppler, tesseract, Ollama/Qwen3-VL, course folders, .course-meta, writable paths, statusline wiring) and optionally auto-repair the permission-free issues. Run when a command won't work or right after cloning.
quiz
Generate N practice problems on a topic. Saves problem MD + hidden answer MD. User solves on paper, uploads answer PDF, then runs /grade. First arg weakmap → auto-load latest weakmap report and target its top weaknesses.