Supported Tools and Features

Rulesync supports both generation and import for All of the major AI coding tools:

Tool	--targets	rules	ignore	mcp	commands	subagents	skills	hooks	permissions	checks
AGENTS.md	agentsmd	✅			🎮	🎮	🎮
AgentsSkills	agentsskills						✅ 🌏
Amp	amp	✅ 🌏		✅ 🌏			✅ 🌏	✅ 🌏	✅ 🌏	✅ 🌏
IBM Bob	bob	✅ 🌏	✅	✅ 🌏	✅ 🌏	✅	✅ 🌏	✅ 🌏	🌏
Claude Code	claudecode	✅ 🌏	✅	✅ 🌏	✅ 🌏	✅ 🌏	✅ 🌏	✅ 🌏	✅ 🌏
Claude Code plugin	claudecode-plugin			✅	✅	✅	✅	✅
CodeBuddy Code	codebuddy	✅ 🌏		✅ 🌏	✅ 🌏	✅ 🌏	✅ 🌏	✅ 🌏	✅ 🌏
Codewhale	codewhale	✅ 🌏		✅ 🌏 🔧	✅ 🌏	✅ 🌏	✅ 🌏	✅ 🌏	✅ 🌏
Codex CLI	codexcli	✅ 🌏		✅ 🌏 🔧	🌏	✅ 🌏	✅ 🌏	✅ 🌏	✅ 🌏
Command Code	commandcode	✅ 🌏		✅ 🌏	✅ 🌏	✅ 🌏	✅ 🌏	✅ 🌏	✅ 🌏
GitHub Copilot	copilot	✅ 🌏		✅	✅	✅ 🌏	✅ 🌏	✅ 🌏	✅
GitHub Copilot CLI	copilotcli	✅ 🌏		✅ 🌏 🔧		✅ 🌏	✅ 🌏	✅ 🌏	✅ 🌏
Crush	crush	✅ 🌏	✅	✅ 🌏 🔧			✅ 🌏	✅ 🌏	✅ 🌏
Goose	goose	✅ 🌏		✅ 🌏	✅ 🌏	✅ 🌏	✅	✅ 🌏	🌏
Hermes Agent	hermesagent	✅	✅	🌏 🔧	🌏	✅ 🌏	🌏	🌏	🌏	✅
Grok CLI	grokcli	✅ 🌏	✅ 🌏	✅ 🌏	✅ 🌏	✅ 🌏	✅ 🌏	✅ 🌏	✅ 🌏
Cursor	cursor	✅	✅	✅ 🌏	✅ 🌏	✅ 🌏	✅ 🌏	✅ 🌏	✅ 🌏	✅
deepagents-cli	deepagents	✅ 🌏		✅ 🌏 🔧		✅ 🌏	✅ 🌏	✅ 🌏	🌏
Factory Droid	factorydroid	✅ 🌏		✅ 🌏 🔧	✅ 🌏	✅ 🌏	✅ 🌏	✅ 🌏	✅ 🌏	✅
OpenCode	opencode	✅ 🌏		✅ 🌏 🔧	✅ 🌏	✅ 🌏	✅ 🌏	✅ 🌏	✅ 🌏
OpenClaw	openclaw	✅ 🌏
MiMo Code	mimocode	✅ 🌏		✅ 🌏 🔧	✅ 🌏	✅ 🌏	✅ 🌏	✅ 🌏	✅ 🌏
GitLab Duo CLI	gitlabduo	✅ 🌏		✅ 🌏	✅ 🌏		✅ 🌏	✅ 🌏		✅
Cline	cline	✅ 🌏	✅	🌏	✅ 🌏	✅ 🌏	✅ 🌏	✅ 🌏	✅
Kilo Code	kilo	✅ 🌏	✅ 🌏	✅ 🌏 🔧	✅ 🌏	✅ 🌏	✅ 🌏	✅ 🌏	✅ 🌏	✅
Kimi Code	kimi-code	✅ 🌏		✅ 🌏 🔧		✅ 🌏	✅ 🌏	🌏	🌏
Kimi Code plugin	kimi-code-plugin	✅			✅	✅	✅
Roo Code ⚠️	roo	✅ 🌏	✅	✅ 🔧	✅ 🌏	✅	✅ 🌏		✅
Zoo Code	zoocode	✅ 🌏	✅	✅ 🔧	✅ 🌏	✅	✅ 🌏		✅
Rovodev (Atlassian)	rovodev	✅ 🌏		✅ 🌏	✅ 🌏	✅ 🌏	✅ 🌏		✅ 🌏	✅
Takt	takt	✅ 🌏		✅ 🌏	✅ 🌏	✅ 🌏	✅ 🌏		✅ 🌏	✅ 🌏
Trae	trae	✅		✅			✅ 🌏
Tabnine CLI ⚠️	tabnine	✅ 🌏	✅	✅ 🌏 🔧	✅ 🌏	✅ 🌏	✅ 🌏	✅ 🌏	✅ 🌏
Continue ⚠️	continue	✅ 🌏	✅ 🌏	✅ 🌏	✅ 🌏		✅ 🌏	✅ 🌏	🌏
Snowflake Cortex Code	cortexcode	✅		🌏		✅ 🌏	✅ 🌏	✅ 🌏
Vibe Code	vibe	✅ 🌏	✅	✅ 🌏 🔧		✅ 🌏	✅ 🌏	✅ 🌏	✅ 🌏
Vibe Code plugin	vibe-plugin			✅		✅	✅	✅
Qoder	qoder	✅ 🌏	✅	✅ 🌏	✅ 🌏	✅ 🌏	✅ 🌏	✅ 🌏	✅ 🌏
Qwen Code	qwencode	✅ 🌏	✅	✅ 🌏 🔧	✅ 🌏	✅ 🌏	✅ 🌏	✅ 🌏	✅ 🌏	✅
Meta Muse Code	musecode	✅		🌏			✅ 🌏
Reasonix	reasonix	✅ 🌏	✅ 🌏	✅ 🌏 🔧	✅ 🌏	✅ 🌏	✅ 🌏	✅ 🌏	✅ 🌏
Codebuff (Freebuff)	codebuff		✅	✅ 🌏
Kiro ⚠️	kiro	✅ 🌏	✅ 🌏	✅ 🌏 🔧	✅	✅	✅	✅	✅
Kiro CLI	kiro-cli	✅ 🌏	✅ 🌏	✅ 🌏 🔧	✅ 🌏	✅ 🌏	✅ 🌏	✅ 🌏	✅
Kiro IDE	kiro-ide	✅ 🌏	✅ 🌏	✅ 🌏 🔧	✅	✅ 🌏	✅ 🌏	✅ 🌏	✅
Letta Code	lettacode		✅			✅ 🌏	✅ 🌏	✅ 🌏	✅ 🌏
Google Antigravity IDE	antigravity-ide	✅ 🌏		✅ 🌏 🔧	✅ 🌏	✅ 🌏	✅ 🌏	✅ 🌏	✅
Google Antigravity CLI	antigravity-cli	✅ 🌏	✅	✅ 🌏 🔧	✅ 🌏	✅ 🌏	✅ 🌏	✅ 🌏	🌏
Google Antigravity plugin	antigravity-plugin	✅		✅ 🔧		✅	✅	✅
JetBrains AI Assistant	aiassistant	✅	✅	✅ 🌏			✅
JetBrains Junie	junie	✅ 🌏	✅	✅ 🌏	✅ 🌏	✅ 🌏	✅ 🌏	🌏	🌏
AugmentCode	augmentcode	✅ 🌏	✅	✅ 🌏	✅ 🌏	✅ 🌏	✅ 🌏	✅ 🌏	✅ 🌏	✅
AugmentCode plugin	augmentcode-plugin	✅		✅	✅	✅	✅	✅
Devin Desktop	devin	✅ 🌏	✅ 🌏	✅ 🌏 🔧	✅ 🌏	✅ 🌏	✅ 🌏	✅ 🌏	✅ 🌏
Devin plugin	devin-plugin	✅		✅ 🔧		✅	✅	✅
Warp	warp	✅ 🌏	✅	✅ 🌏	✅ 🌏		✅ 🌏		🌏
Warp Agent CLI	warpcli			🌏					🌏
Replit	replit	✅					✅ 🌏
Pi Coding Agent	pi	✅ 🌏		✅ 🌏	✅ 🌏		✅ 🌏	✅ 🌏	✅ 🌏
oh-my-pi	omp	✅ 🌏		✅ 🌏	✅ 🌏	✅ 🌏	✅ 🌏	✅ 🌏	✅ 🌏
Zed	zed	✅ 🌏	✅ 🌏	✅ 🌏			✅ 🌏		✅ 🌏
ZCode (Z.ai)	zcode	✅ 🌏		✅ 🌏	✅ 🌏	✅ 🌏	✅ 🌏	✅ 🌏
ZCode plugin	zcode-plugin			✅	✅	✅	✅	✅
Pool (Poolside)	pool	✅ 🌏		✅ 🌏 🔧		✅ 🌏	✅ 🌏	✅ 🌏	✅ 🌏
DeepSeek Harness	dsh	✅ 🌏		🌏			✅ 🌏

• ✅: Supports project mode
• 🌏: Supports global mode
• 🎮: Supports simulated commands/subagents/skills (Project mode only)
• 🔧: Supports MCP tool config (enabledTools/disabledTools)
• ⚠️: Deprecated — still supported, but see the note below

Hermes Agent compatibility

The hermesagent target is validated against Hermes Agent v0.21.3 (release v2026.9.14). The supported contract covers project rules, ignore patterns, subagents, and checks, plus global MCP servers, commands, subagents, skills, hooks, and permissions. Generation, --check, and import round-trips are covered for both advertised scopes.

Rulesync honors Hermes profiles through HERMES_HOME. When it is set, its value is the profile root itself: global configuration is read and written directly under $HERMES_HOME (config.yaml, skills/, plugins/, and rulesync/), without appending .hermes. When it is unset, Rulesync follows Hermes's own platform default: ~/.hermes everywhere except Windows, where it is %LOCALAPPDATA%\hermes. Because HERMES_HOME names where Hermes itself reads the profile, it also takes precedence over --output-roots in global scope. Project-scoped paths remain rooted in the project.

Changing which profile root Rulesync resolves strands whatever it generated under the previous one. --delete reconciles only the root resolved for the current run, so files under a root it no longer resolves are invisible to it and must be removed by hand once you are sure Hermes no longer reads them. This applies whenever you set or change HERMES_HOME, and to two upgrades that moved the resolved root on their own: before v16.0.0 global files went to ~/.hermes even when HERMES_HOME was set, and before v16.2.0 they went there on Windows too, rather than to %LOCALAPPDATA%\hermes.

Project plugins are registered by adding their names to $HERMES_HOME/config.yaml, but Rulesync does not persist Hermes's global project-plugin trust gate. Run Hermes from a trusted project root with HERMES_ENABLE_PROJECT_PLUGINS=true for an explicit, session-scoped opt-in. A future Hermes release that changes its loaders, schemas, or plugin API requires a new compatibility validation.

Deprecation notes

• Google Antigravity (antigravity-ide / antigravity-cli) — Antigravity 2.0 splits into two products: the desktop antigravity-ide and the antigravity-cli (agy). As of Antigravity 2.0 the IDE reads its global MCP config and skills from the shared ~/.gemini/config/ tree — ~/.gemini/config/mcp_config.json and ~/.gemini/config/skills/, matching the current MCP and Skills docs. The antigravity-cli global MCP config also lives in the shared ~/.gemini/config/mcp_config.json, while the CLI keeps its own global skills tree at ~/.gemini/antigravity-cli/skills/. Both targets also intentionally share the global rule file ~/.gemini/GEMINI.md and the global hooks file ~/.gemini/config/hooks.json — enabling both targets in --global mode writes those shared files once. For project-scope rules, both antigravity-ide and antigravity-cli emit the root rule as a plain cross-tool AGENTS.md at the project root (the Gemini-lineage discovery order is AGENTS.md, CONTEXT.md, GEMINI.md; the IDE has read AGENTS.md since v1.20.3) and non-root rules under .agents/rules/ (both add trigger frontmatter to non-root rules, because Antigravity silently discards any .md file under a rules/ directory that lacks a valid trigger, so the two targets write identical files there). In --global mode, both targets keep the root rule in ~/.gemini/GEMINI.md and write each non-root rule to its own file in ~/.gemini/config/rules/ with trigger frontmatter (always_on unless the rule's globs or its antigravity.trigger say otherwise), because Antigravity truncates every rule file at 24,000 bytes and loads that directory as separate global rules (see the Antigravity rules docs). A nested rule is flattened to a top-level name there (frontend/style.md → frontend-style.md), since Antigravity reads only that directory's top level. The two targets write identical files there. The directory holds global rules created outside rulesync too, so generate --delete never removes files there: after deleting or renaming a rulesync rule, remove its old file by hand. The per-file limit is not the only cap: always-on and global rules also share a 20,000-token budget, over which rules are demoted to path pointers. For commands, Antigravity retires workflows on October 19, 2026 and documents skills as their replacement (a skill is invoked as /<skill-name> exactly like a workflow, and wins over a workflow of the same name), so both targets emit each command as a skill: .agents/skills/<name>/SKILL.md in the project, and ~/.gemini/config/skills/<name>/SKILL.md (IDE) or ~/.gemini/antigravity-cli/skills/<name>/SKILL.md (CLI) in --global mode — the same directories the skills feature writes. <name> is the command's trigger (antigravity.trigger, then a root trigger, then a trigger: /name line in the body, defaulting to the filename), and the skill gets name/description frontmatter. A rulesync skill with the same name that is written to the same skills directory (it targets * or a tool sharing that directory) takes precedence: the command is skipped with a warning, whichever targets and features a given run includes, so a partial run never overwrites the skill. The commands feature does not sweep the shared skills tree, so the skill of a deleted command is removed by generate --delete with the skills feature enabled. rulesync import still reads legacy workflows from .agents/workflows/ (project) and ~/.gemini/antigravity/global_workflows/ / ~/.gemini/antigravity-cli/global_workflows/ (global), and generate --delete removes every *.md file left in those workflow directories, including hand-written ones (import them first to keep them).
• Continue (continue) — Continue is end of life: Continue joined Cursor on 2026-06-18, the continuedev/continue repository is read-only with a final v2.0.0 VS Code release (2026-06-19) and its last commit on 2026-07-21, and the in-product deprecation banner expired on 2026-09-09. The continue target stays supported as a frozen-compatibility target: every adapter was verified against that final source tree (commit 5522c6f4, which the source links in the code and docs are pinned to), and nothing about the file formats will move again, so existing output keeps working for anyone still running the last release. Do not expect new Continue capabilities to be tracked; new projects should pick an actively maintained tool instead.
• Kiro (kiro) — Kiro ships as two products with diverging config formats: the Kiro IDE reads Markdown subagents (.kiro/agents/*.md) and structured JSON hooks (.kiro/hooks/*.json, format { "version": "v1", "hooks": [ ... ] }), while the Kiro CLI reads JSON agent-config subagents (.kiro/agents/*.json). A single target cannot emit both subagent shapes faithfully, so kiro is split into kiro-cli and kiro-ide. The legacy kiro target is kept as a deprecated alias (its current mixed output is unchanged for backward compatibility). Shared surfaces (steering rules with inclusion, .kiro/settings/mcp.json, .kiro/prompts/ commands, .kiro/skills/, .kiroignore, permissions) are identical between the two; they differ only in subagents (.md vs .json). Hooks are the same for both: a single .kiro/hooks/rulesync.json (whose hooks array holds every generated hook) in both project (.kiro/hooks/) and global (~/.kiro/hooks/) scope, mapping canonical lifecycle events to Kiro's PascalCase triggers (SessionStart, UserPromptSubmit, PreToolUse, PostToolUse, Stop) and supporting both agent (prompt) and command actions. Kiro CLI 3.0 migrated to that format and no longer reads the embedded agent hooks in .kiro/agents/default.json, which only the deprecated kiro alias still writes (including cacheTtl ⇄ cache_ttl_seconds). Global skills (~/.kiro/skills/), global ignore (~/.kiro/settings/kiroignore), and global Kiro IDE subagents (~/.kiro/agents/) are also supported, as are global Kiro CLI commands (~/.kiro/prompts/) and subagents (~/.kiro/agents/). Kiro's shared MCP file preserves per-server disabledTools.
• Roo Code (roo) — Roo Code is end of life: its final release was v3.54.0 (2026-05-15) and the Roo-Code repository is archived, so nothing about the target will move again. New projects should target zoocode instead — Zoo Code is the community continuation named by the Roo shutdown notice, and it continues Roo's release numbering. The roo target stays supported because Zoo Code still reads the .roo/ project tree and ~/.roo global tree verbatim, so existing roo output keeps working; what it no longer tracks is anything Zoo Code added after the fork. The two targets write the same files, so enable one per project rather than both — see the Zoo Code note in File formats for the fail-open hazard a --targets roo generate creates in a shared .roomodes. Permissions are the one exception: the two lineages spell the command allow/deny lists differently (roo-cline.* vs zoo-code.*), so both pairs coexist in one .vscode/settings.json and neither target touches the other's — see the Roo Code paragraph in File formats.
• Tabnine CLI (tabnine) — The file-based Tabnine CLI this target models is in maintenance mode: as of Tabnine 6.6 (2026-09-15) it is "Tabnine CLI (Legacy)", installed as tabnine-legacy once the new CLI is present, and it receives only critical updates through 2026-12-31, after which Tabnine deprecates it fully. The current tabnine command is Tabnine's distribution of OpenCode, configured with standard OpenCode files — so for the new CLI use --targets opencode (AGENTS.md, opencode.json, .opencode/{commands,agents,skills}, ~/.config/opencode/**); its migration helper (/migrate, /migrate-context) carries MCP servers, skills, agents, commands and TABNINE.md over. The new CLI adds ~/.tabnine/opencode/config/ as an extra config root but still reads the standard OpenCode roots — project opencode.json(c) and .opencode/, and ~/.config/opencode/** (configuration docs) — so --targets opencode output (including --global) is picked up; keys set in Tabnine's own opencode.json there are layered last and win over the same keys in the generated one. A global AGENTS.md is not read from ~/.tabnine/opencode/config/; it comes from ~/.config/opencode/AGENTS.md, where Rulesync writes it. The tabnine target stays supported for the legacy CLI (0.36.0 as of Tabnine 6.6.5) until then; whether it is removed in a later major is decided separately.