Borrowing it
Nothing to install: this file belongs to Beowolve/SnapFX. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.
curl -O https://raw.githubusercontent.com/Beowolve/SnapFX/main/AGENTS.mdgit clone --depth 1 https://github.com/Beowolve/SnapFXWrote 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.
[](https://agentmods.dev/instructions/beowolve/snapfx/agents-md)<a href="https://agentmods.dev/instructions/beowolve/snapfx/agents-md"><img src="https://agentmods.dev/badge/instructions/beowolve/snapfx/agents-md/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.
<a href="https://agentmods.dev/instructions/beowolve/snapfx/agents-md"><img src="https://agentmods.dev/badge/instructions/beowolve/snapfx/agents-md.svg" alt="Reviewed on agentmods" width="80" height="20"></a>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.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.02010 | $0.02010 |
| Opus 5.5 | $0.00804 | $0.00804 |
| Sonnet 5.5 | $0.00402 | $0.00402 |
| Haiku 4.5 | $0.00201 | $0.00201 |
Grade A, and why
SnapFX 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 2mo 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 — 67 lines — stays where its author put it; the contents beside it link to each section on GitHub.
AI Agent Definitions
These instructions are shared for all AI agents working on this workspace.
Core Workflow
- Always add or update unit tests for new behavior, bug fixes, or refactors.
- For bug fixes, first add or update a test that reproduces the issue and verify it fails before implementing the fix.
- Exception: if a bug is only reproducible via user interaction and cannot be automated yet, document manual reproduction/verification steps and add or improve logging so the failure path is observable.
- Always run unit tests after code changes when possible (
./gradlew test). - Always review STATUS.md, DONE.md, and ROADMAP.md after changes and update each file only when the change is relevant to that file's scope (
STATUS= current state/open issues,DONE= completed milestones/capabilities,ROADMAP= planned/proposed work). - Keep CHANGELOG.md up to date for release-visible changes (
Unreleasedduring development, grouped by tags for releases). - Update other
*.mdfiles if the change affects their content or accuracy. - Keep changes scoped and avoid unrelated edits.
- After successful changes, prepare a small, focused commit (stage relevant files and propose a commit message).
- Always
git addnew files when preparing a commit. - Before creating a commit, always show the proposed commit message to the user and get confirmation.
- Commit-message focus rule: for feature commits, describe the delivered user-visible/technical feature scope (not temporary sprint/phase labels), and do not list intermediate bug fixes encountered during feature development unless the commit is primarily a bug-fix commit.
- Keep commits minimally mixed: one logical topic/fix per commit; avoid bundling unrelated changes.
- Default to one commit per fix/feature so changes stay clearly separated and release notes remain precise.
- For fix commits, use a multi-line commit message body with at least one explanation line per fix.
- Use git-cliff-compatible commit prefixes in the subject line (
feat:,fix:,docs:,test:,refactor:,perf:,ci:,build:,chore:); domain prefixes likefloating:,dnd:,ui:,demo:are also allowed when they map to cliff groups. - If a fixed collaboration rule is agreed with the user, add it to this AGENTS.md immediately.
- Treat AGENTS.md as the source of truth for all persistent collaboration rules in this workspace.
- API-to-settings parity rule: every SnapFX API function must be represented in the MainDemo Settings tab to keep manual verification accessible.
- Design-decision documentation rule: every significant design/architecture decision must be documented in JavaDoc (API/algorithm intent), in ARCHITECTURE.md (system behavior), and as an ADR file under
docs/adr/. - JavaDoc completeness rule: for every new or changed public/protected API element (class, method, enum constant, public field), add/update full JavaDoc immediately in the same change (
@param,@return,@throwswhere applicable), run./gradlew :snapfx-core:javadoc --rerun-tasks, and do not defer missing JavaDoc cleanup to later follow-up work. - As soon as a focused, sensible commit is ready, proactively show the proposed commit message without waiting for the user to ask.
- If
MainDemovisuals change (layout, window composition, look and feel, styling, icons, controls), always run./scripts/update-main-demo-preview.ps1and include the updateddocs/images/main-demo.png. - Versioning is controlled by
gradle-jgitverinbuild.gradle.kts; do not reintroduce custom version calculators in the build script. - Use tag-driven releases (
vX.Y.Z) as the only release source of truth; do not create tags per commit. - Maintainability-refactor tagging rule: when roughly half of the planned pre-
v1.0.0maintainability refactor scope is fully completed, create tagv0.7.0; when that refactor scope is fully completed, move project/release baseline tov0.8.0. - Naming/package convention rule: except for
SnapFXandBuildInfo, framework classes should use theDockprefix; extracted/refactored implementation types must be placed in thematic packages (for exampleorg.snapfx.shortcuts) instead of a catch-allorg.snapfx.internal.snapfx, and related types for one concern should stay co-located in that thematic package. - Utility packaging rule: place extracted classes in
org.snapfx.utilonly when they are true cross-domain utilities; otherwise keep them in the most fitting thematic package for their concern. - Branch workflow policy: while base implementation is ongoing, work directly on
main; oncedevelopexists, integrate features intodevelopand cut release tags frommainafter merge. - Pre-release compatibility policy: until the first public Maven release, prefer best technical fixes over backward compatibility; breaking API changes are acceptable when they improve correctness/maintainability. After the first Maven release, public API compatibility becomes mandatory unless explicitly planned as a breaking release.
- Status documentation consistency rule: in STATUS.md, DONE.md, and ROADMAP.md, status bullets must always include a status icon prefix (
✅,🚧,📋,💡,❌,⚠️), never plain- Added/Updated/Completed/.... - Change-history rule: STATUS.md should stay current-state focused; versioned historical details belong in CHANGELOG.md.
- Status/DONE readability rule: keep STATUS.md as a concise current-state snapshot (no changelog/commit-style bullet history), keep DONE.md at milestone level only, and keep per-test-case breakdowns/count churn out of both files (details belong in CI output, test reports, or CHANGELOG.md when release-relevant).
- Refactor-progress documentation rule: during ongoing multi-step refactor tracks, do not log each implementation slice in STATUS.md or ROADMAP.md; keep both files high-level and update them at milestone/phase boundaries or when the tracked refactor scope is completed.
- Changelog style rule: write CHANGELOG.md entries in concise release-note style (grouped outcomes), not as step-by-step internal worklog/task-sequencing bullets.
- Changelog category rule: structure
Unreleased(and new release sections going forward) using git-cliff-style categories in this order (include applicable sections only):Features,Fixes,UI and Interaction,Refactoring,Tests,Documentation,Build and Tooling,Other Changes. - Changelog scope rule: do not add AGENTS.md-only collaboration/workflow rule changes to CHANGELOG.md; keep those developer-internal rule updates out of release notes.
- Markdown encoding rule: preserve UTF-8 and avoid shell text-rewrite commands that can alter Unicode; prefer
apply_patchfor markdown edits. - Temporary-file rule: markdown files matching
*_temp.mdare local working notes only, must not be committed, and must not be referenced from repository status/history docs. - Diff minimization rule: for all file changes, prefer minimal targeted edits and avoid full-file rewrites when smaller diffs are sufficient; only rewrite entire files when explicitly requested or technically required.
- Logic deduplication rule: when new behavior overlaps existing behavior (for example restore/fallback flows), reuse or extract the existing implementation path instead of re-implementing parallel logic.
- Unicode escape readability rule: when icon constants use
\u...notation in code, add an inline comment showing the rendered icon.
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.
- 2mo ago First seen · 67 lines · 2,010 tokens per session scan A a871ff1e9520
SnapFX AGENTS.md is an instructions file published in the GitHub repository Beowolve/SnapFX (23 stars, last pushed 6mo ago), licensed MIT. It adds 2,010 tokens to every session, about $0.0080 per session on Opus 5.5. 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.
SnapFX AGENTS.md compared with instructions from other repositories
| Mod | Type | Stars | Last push | Licence | Scan | Tokens | Works with |
|---|---|---|---|---|---|---|---|
| SnapFX AGENTS.mdBeowolve/SnapFX | Instructions file | 23 | 6mo ago | MIT | A | 2,010 | Codex, OpenCode |
| vscode buildNext.instructions.mdmicrosoft/vscode | Instructions file | 195k | yesterday | MIT | A | 9,719 | GitHub Copilot |
| next.js AGENTS.mdvercel/next.js | Instructions file | 143k | yesterday | MIT | A | 7,461 | Codex, OpenCode |
| langchain AGENTS.mdlangchain-ai/langchain | Instructions file | 149k | yesterday | MIT | A | 4,680 | Codex, OpenCode |
| vscode oss-third-party-notices.instructions.mdmicrosoft/vscode | Instructions file | 195k | yesterday | MIT | A | 5,516 | GitHub Copilot |
| vscode copilot-instructions.mdmicrosoft/vscode | Instructions file | 195k | yesterday | MIT | A | 3,093 | GitHub Copilot |
| deepseek-harness AGENTS.mddeepseek-ai/deepseek-harness | Instructions file | 274k | 2d ago | MIT | A | 4,313 | Codex, OpenCode |