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 skills/ai-4-phi/phillit/setupnpx skills add AI-4-Phi/PhilLit --skill setupgit clone --depth 1 https://github.com/AI-4-Phi/PhilLitWhat 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 | $0.00068 | $0.01346 |
| Opus 5 | $0.00034 | $0.00673 |
| Sonnet 5 | $0.00014 | $0.00269 |
| Haiku 4.5 | $0.00007 | $0.00135 |
Grade A, and why
setup scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 2d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 81 lines — stays where its author put it; the contents beside it link to each section on GitHub.
PhilLit workspace setup
Run this once in the directory where the user will create reviews.
Communication style: keep every user-facing message to a sentence or two of plain, non-technical language. Do everything you can on the user's behalf; ask only for what only they can provide.
- Open with what this is. Before running anything, tell the user in 1–2 sentences, e.g.: "Let's set up PhilLit in this folder — a one-time step. I'll create the configuration for you; I'll only need your email address and one free API key."
- Check the environment. Run
[ -x "${PHILLIT_ROOT:-}/bin/phillit-run" ] && echo ok || echo missing,command -v uv, andcommand -v jq.- If the first check prints
missing, the plugin's environment isn't loaded in this session (PHILLIT_ROOTunset, empty, or stale — typically the plugin was installed or updated mid-session; the bridge only runs at session start), so every later step would fail with a cryptic "No such file or directory". Tell the user: "One quick step: please restart Claude Code in this folder, then run /phillit:setup again." and stop. - If uv or jq is absent, give a one-line install instruction (uv:
https://docs.astral.sh/uv/getting-started/installation/ ; jq:
brew install jq/apt install jq/choco install jq) and stop —jqis required by the BibTeX-validation hook and is otherwise a silent clean-install failure.
- If the first check prints
- Preview and ask consent. Run:
bash "$PHILLIT_ROOT/bin/phillit-run" skills/setup/scripts/setup_workspace.py --plugin-root "$PHILLIT_ROOT" --dry-runThen summarize what will happen in 2–3 short bullets — do not dump the settings JSON (show it only if the user asks for details):- Creates a
.phillitfolder marker and, for any API keys not already set in the environment, a.envfile to hold them (keys found in the environment are used from there — never copied into a file) - Lets Claude run PhilLit's research tools in this folder without asking permission each time; deleting files and system-level commands still require approval
- Nothing is sent or published anywhere Ask whether to proceed.
- Creates a
- Apply. On approval, run the same command without
--dry-run. Keys already set in the user's environment stay there — the script never copies their values into.env; it prints which ones it found (and skips creating.envwhen it found them all). If it found some, tell the user (e.g. "Your Brave key was already set up — I'm using it from your environment."). - Collect the rest. Ask only for what is still missing, and be explicit about
required vs. optional. Edit
.envyourself with the answers (uncomment the key's# KEY=line and add the value) — don't make the user open a file. Keys the user skips stay commented: an active emptyKEY=line would override that key with "" if the user ever exports it in their shell.- Required:
CROSSREF_MAILTO— just the user's email address; ask for it and fillOPENALEX_EMAILwith the same addressBRAVE_API_KEY— the one thing the user must obtain themselves: a free key from https://brave.com/search/api/
- Optional (skipping is fine):
S2_API_KEY(https://www.semanticscholar.org/product/api) andCORE_API_KEY(https://core.ac.uk/services/api) improve search coverage and speed.
- Required:
- Verify. Run
bash "$PHILLIT_ROOT/bin/phillit-run" skills/philosophy-research/scripts/check_setup.pyand report in one line: "Setup complete — ask me for a literature review whenever you're ready", or exactly what is still missing and how to fix it. - Trust check. Settings changes reload live, but Claude Code ignores the new
allowrules until the workspace is trusted. Check (prefer the$PWDentry; fall back to the physical path only when$PWDhas no entry at all — under a symlinked cwd, e.g. OneDrive or/var→/private/var,.claude.jsonmay key either form):CFG="${CLAUDE_CONFIG_DIR:+$CLAUDE_CONFIG_DIR/.claude.json}" jq -r --arg d "$PWD" --arg p "$(pwd -P)" \ '((.projects[$d] // .projects[$p] // {}).hasTrustDialogAccepted // false)' \ "${CFG:-$HOME/.claude.json}"- If
true(or the check itself fails): say nothing — permissions are already live. - If
false: end the setup with this, prominently, as the final line: "One last step: restart Claude Code in this folder and choose 'Yes, I trust this folder' when asked — PhilLit's permissions only take effect after that."
- If
- Updates tip. Close with one sentence: "PhilLit doesn't update automatically — to
change that, open
/plugin, go to Marketplaces, and turn on auto-update for phillit (or run/plugin update phillit@phillitwhenever you want the latest)."
What ships with it
1 file 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.
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.
- 2d ago First seen · 81 lines · 68 tokens per session scan A 801e7735cad1
setup is a skill published in the GitHub repository AI-4-Phi/PhilLit (41 stars, last pushed 2d ago), licensed Apache-2.0. It adds 68 tokens to every session and 1,346 once invoked, about $0.0003 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 skills, from other repositories
life-meaning
当用户问「活着的意义是什么/找不到方向/虚无感/该为什么努力」时调用。 核心理念: 生命没有外在赋予的意义(宇宙尺度一切会消失); 意义=个人创造+愿意为之受苦; 拥抱死亡是意义的入口。 不适用于: 急性自杀风险(先紧急求助)。 Triggers: 生命的意义/虚无/找不到意义/为什么活着/meaning/purpose/nihilism.
rational-buddhism
当用户问某主张/玄学/灵修该不该信、想建立自己的验证标准、或寻找科学与精神的结合点时调用。 核心理念: 尝试一切, 亲自测试, 保持怀疑, 保留有用的, 舍弃没用的; 不可证伪的主张不进入基本真理; 冥想/接纳等内在技术可保留。 不适用于: 需要信仰安慰的临终/哀伤场景(尊重个体选择)。 Triggers: 该不该信/玄学/灵修/冥想/可证伪/验证/理性/rational buddhism/verify/belief.
roundtable
Multi-perspective roundtable discussion — a truth-seeking moderator plus a small set of representative real figures explore one topic through disciplined debate. Use for 圆桌讨论, roundtable, 多人物观点碰撞, 让几位思想家讨论, 结构化辩论探索. Not for ordinary argumentative writing, pros/cons lists, or single-speaker explanation.
academic-paper-composer
Systematic writing framework for philosophy and interdisciplinary academic papers from optimized outline to submission-ready manuscript. Use when users want to: (1) write a paper from a detailed outline, (2) ensure quality control during writing, (3) maintain consistency across chapters, (4) prepare a submission-ready…
academic-paper-strategist
Systematic strategic planning framework for philosophy and interdisciplinary academic papers targeting preprint platforms (PhilArchive, arXiv, PhilSci-Archive). Use when users want to: (1) plan a paper on a specific topic, (2) identify research gaps and assess originality, (3) develop optimized paper outlines, (4)…
worldview-synthesis
This skill should be used when someone wants to articulate, explore, or document their personal worldview, values, or philosophy. Triggers on "articulate my values", "figure out what I believe", "document my philosophy", "write a manifesto", "define my leadership philosophy", "explore my beliefs". Surfaces beliefs…