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 instructions/ripemangobox/bite/agents-mdgit clone --depth 1 https://github.com/RipeMangoBox/BITEWhat 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 | $0.00675 | $0.00675 |
| Opus 5 | $0.00338 | $0.00338 |
| Sonnet 5 | $0.00135 | $0.00135 |
| Haiku 4.5 | $0.00068 | $0.00068 |
Grade A, and why
BITE AGENTS.md 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 — 69 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Agent Guide
Public agent-facing usage starts from README.md and
.claude/skills/README.md. Internal deployment notes
stay under _private/.
BITE is a local-first research workflow for structured paper analysis. Its goal is to turn academic PDFs into machine-actionable evidence notes, retrieval indexes, and downstream idea or review artifacts. PaperBite acts as the upstream public evidence layer: it provides reusable structured paper assets, while BITE focuses on local analysis, retrieval, comparison, idea generation, and research decision support.
Working Surface
obsidian-vault/paperPDFs/stores source PDFs.obsidian-vault/analysis/stores structured local analysis notes.obsidian-vault/index/stores generated indexes and Obsidian navigation.obsidian-vault/ideas/stores local idea, focus, and review notes.
Local Pipeline
collect candidate papers / import local PDFs
-> download when needed
-> integrated analysis chain
(MinerU parse/reuse -> structured analysis -> vault export)
-> optional index refresh
-> query / ideate / focus / review / export
Branch Sync Policy
This branch is the canonical place for changes that should later appear in the
public main worktree. Make shared README, public docs, public assets, and
public workflow updates here first, then sync them to main after review.
Keep branch-specific and local-development material out of main:
_private/local notes, archives, deployment notes, and operation history- Obsidian workspace/runtime state
- provider-specific defaults or private model choices
- internal branch/worktree coordination notes
AGENTS.md itself is branch-specific. Do not blindly sync this file to
main; the public main copy should contain only project background and
public agent-facing guidance.
Rules
- Treat the local vault paths above as the current working surface.
- Write only through the skill that owns the target output path.
- Analysis language defaults to
zhunless the request overrides it. - Pipeline steps are idempotent; already-completed steps should be skipped.
- Planned analysis batches must declare goal, source, selection rule, budget, and output target before agents run.
- Agents must preserve source anchors in notes, logs, and generated outputs.
- Reports and profiles must be generated from available evidence, not new unsupported claims.
- When prose needs to mention Markdown or Obsidian reserved characters such as
*,[,],|, or#, escape them with backslashes or wrap them in inline code so reading view does not reinterpret the text. - In Markdown tables, aliased Obsidian wikilinks are allowed only when the
alias separator is escaped as
[[full/path\|abbr]]; an unescaped|splits table columns. Outside tables, normal[[full/path|abbr]]links are fine. - Generated exports, snapshots, backups, local storage, and symlinks stay out of Git.
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 · 69 lines · 675 tokens per session scan A 5e7018336e83
BITE AGENTS.md is an instructions file published in the GitHub repository RipeMangoBox/BITE (60 stars, last pushed 19d ago), licensed MIT. It adds 675 tokens to every session, about $0.0034 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 instructions, from other repositories
K-Ops GEMINI.md
Gemini CLI instructions for arnaudgelas/K-Ops, covering gemini.md, gemini-specific notes, mission, operating rules and default workflow.
K-Ops AGENTS.md
AGENTS.md instructions for arnaudgelas/K-Ops, covering agents.md, roles, source ingestor, wiki compiler and lint + heal.
K-Ops CLAUDE.md
Claude Code instructions for arnaudgelas/K-Ops, covering claude.md, mission, operating rules, default workflow and research runs.
obsidian-mind CLAUDE.md
Claude Code instructions for breferrari/obsidian-mind, covering obsidian mind, skills & capabilities, custom slash commands, vault structure and obsidian cli.
llm-wiki-agent AGENTS.md
AGENTS.md instructions for SamurAIGPT/llm-wiki-agent, covering llm wiki agent — schema & workflow instructions, how to use, directory layout, page format and ingest workflow.
llm-wiki-agent CLAUDE.md
Claude Code instructions for SamurAIGPT/llm-wiki-agent, covering llm wiki agent — schema & workflow instructions, slash commands (claude code), directory layout, page format and ingest workflow.