| Dimension | Shared | Only in A | Only in B | Overlap |
|---|---|---|---|---|
| Sections | 0 | 3 | 5 | 0% |
| Commands | 0 | 3 | 0 | 0% |
| Section tags | 1 | 2 | 1 | 25% |
What each file covers
Sections
0 shared · 3 only in A · 5 only in B- − Agent Skills
- − Layout
- − Editing
- + .github Quick Reference
- + Key Files
- + Workflow Naming
- + Common Tasks
- + Reference
Commands
0 shared · 3 only in A · 0 only in B- − git config core.symlinks true
- − pnpm sync:skill-links
- − pnpm check:skill-links
Section tags
1 shared · 2 only in A · 1 only in B- − setup
- − architecture
- + code-style
- agent-behaviour
Line diff
n8n-io/n8n · .agents/skills/AGENTS.md
@@ −1 @@
1# Agent Skills
2
3Shared n8n skills live in `.agents/skills`. These are the canonical source for
4skills that should work across Claude Code, OpenCode, and future agent harnesses.
5
6## Layout
7
8- Shared skills: `.agents/skills/<name>/SKILL.md`
9- Claude-specific skills and overrides: `.claude/plugins/n8n/skills/<name>/`
10- OpenCode-specific skills and overrides: `.opencode/skills/<name>/`
11
12Claude plugin skills should usually be symlinks back to `.agents/skills`.
13OpenCode discovers `.agents/skills` directly, so `.opencode/skills` should only
14contain OpenCode-specific real-directory overrides.
15
16Real directories in a harness path are treated as intentional overrides and are
17preserved by the sync script. The sync script only prunes symlinks that point
18back into `.agents/skills` (stale shared-skill links); hand-placed symlinks
19pointing elsewhere are left untouched.
20
21The Claude plugin links are git symlinks, so a checkout needs symlink support.
22On Windows, enable it (`git config core.symlinks true`, plus Developer Mode or
23WSL) before checking out — otherwise git writes the links as plain text stubs
24and Claude Code fails to load those skills. `--check` reports stubs with an
25actionable error. (OpenCode reads `.agents/skills` directly and is unaffected.)
26
27## Editing
28
29- Edit shared skills under `.agents/skills`, not through symlinked copies.
30- Only put a skill in `.agents/skills` if it can work across supported agent
31 harnesses. Avoid hardcoded harness tool names, commands, or UI flows unless
32 the skill includes a clear availability check and fallback.
33- Keep harness-specific commands out of shared skills unless guarded by clear
34 tool availability checks.
35- Put harness-specific workflows, such as MCP setup commands, in real
36 directories under the matching harness path.
37- Run `pnpm sync:skill-links` after adding or removing shared skills to update
38 Claude plugin symlinks.
39- Run `pnpm check:skill-links` before submitting changes.
40
n8n-io/n8n · .github/CLAUDE.md
@@ +1 @@
1@../AGENTS.md
2
3## .github Quick Reference
4
5This folder contains n8n's GitHub Actions infrastructure.
6
7### Key Files
8
9| File/Folder | Purpose |
10|-------------|---------|
11| `WORKFLOWS.md` | Complete CI/CD documentation |
12| `DEVELOPING_V3.md` | How to develop v3 features (master + 3.x branch model, opt-in flags) |
13| `workflows/` | GitHub Actions workflows |
14| `actions/` | Reusable composite actions |
15| `scripts/` | Release & Docker automation |
16| `CODEOWNERS` | Team review ownership |
17
18### Workflow Naming
19
20| Prefix | Purpose |
21|--------|---------|
22| `test-` | Testing (unit, E2E, visual) |
23| `ci-` | Continuous integration |
24| `util-` | Utilities (notifications) |
25| `build-` | Build processes |
26| `release-` | Release automation |
27| `sec-` | Security scanning |
28
29Reusable workflows: add `-reusable` or `-callable` suffix.
30
31### Common Tasks
32
33**Add workflow:** Create in `workflows/`, document in `WORKFLOWS.md`.
34Always declare a least-privilege top-level `permissions:` block (usually
35`contents: read`) — without one the workflow runs with the repo's broad
36default token and review flags it. Jobs needing more override at job level;
37a job calling a reusable workflow must grant at least what that workflow
38declares.
39
40**Add script:** Create `.mjs` in `scripts/`, document in `WORKFLOWS.md`
41
42### Reference
43
44See `WORKFLOWS.md` for:
45- Architecture diagrams
46- Workflow call graph
47- Scheduled jobs & triggers
48- Runners & secrets
49
@@ −1 +1 @@
1−# Agent Skills
1+@../AGENTS.md
22
3−Shared n8n skills live in `.agents/skills`. These are the canonical source for
4−skills that should work across Claude Code, OpenCode, and future agent harnesses.
3+## .github Quick Reference
54
6−## Layout
5+This folder contains n8n's GitHub Actions infrastructure.
76
8−- Shared skills: `.agents/skills/<name>/SKILL.md`
9−- Claude-specific skills and overrides: `.claude/plugins/n8n/skills/<name>/`
10−- OpenCode-specific skills and overrides: `.opencode/skills/<name>/`
7+### Key Files
118
12−Claude plugin skills should usually be symlinks back to `.agents/skills`.
13−OpenCode discovers `.agents/skills` directly, so `.opencode/skills` should only
14−contain OpenCode-specific real-directory overrides.
9+| File/Folder | Purpose |
10+|-------------|---------|
11+| `WORKFLOWS.md` | Complete CI/CD documentation |
12+| `DEVELOPING_V3.md` | How to develop v3 features (master + 3.x branch model, opt-in flags) |
13+| `workflows/` | GitHub Actions workflows |
14+| `actions/` | Reusable composite actions |
15+| `scripts/` | Release & Docker automation |
16+| `CODEOWNERS` | Team review ownership |
1517
16−Real directories in a harness path are treated as intentional overrides and are
17−preserved by the sync script. The sync script only prunes symlinks that point
18−back into `.agents/skills` (stale shared-skill links); hand-placed symlinks
19−pointing elsewhere are left untouched.
18+### Workflow Naming
2019
21−The Claude plugin links are git symlinks, so a checkout needs symlink support.
22−On Windows, enable it (`git config core.symlinks true`, plus Developer Mode or
23−WSL) before checking out — otherwise git writes the links as plain text stubs
24−and Claude Code fails to load those skills. `--check` reports stubs with an
25−actionable error. (OpenCode reads `.agents/skills` directly and is unaffected.)
20+| Prefix | Purpose |
21+|--------|---------|
22+| `test-` | Testing (unit, E2E, visual) |
23+| `ci-` | Continuous integration |
24+| `util-` | Utilities (notifications) |
25+| `build-` | Build processes |
26+| `release-` | Release automation |
27+| `sec-` | Security scanning |
2628
27−## Editing
29+Reusable workflows: add `-reusable` or `-callable` suffix.
2830
29−- Edit shared skills under `.agents/skills`, not through symlinked copies.
30−- Only put a skill in `.agents/skills` if it can work across supported agent
31− harnesses. Avoid hardcoded harness tool names, commands, or UI flows unless
32− the skill includes a clear availability check and fallback.
33−- Keep harness-specific commands out of shared skills unless guarded by clear
34− tool availability checks.
35−- Put harness-specific workflows, such as MCP setup commands, in real
36− directories under the matching harness path.
37−- Run `pnpm sync:skill-links` after adding or removing shared skills to update
38− Claude plugin symlinks.
39−- Run `pnpm check:skill-links` before submitting changes.
31+### Common Tasks
32+
33+**Add workflow:** Create in `workflows/`, document in `WORKFLOWS.md`.
34+Always declare a least-privilege top-level `permissions:` block (usually
35+`contents: read`) — without one the workflow runs with the repo's broad
36+default token and review flags it. Jobs needing more override at job level;
37+a job calling a reusable workflow must grant at least what that workflow
38+declares.
39+
40+**Add script:** Create `.mjs` in `scripts/`, document in `WORKFLOWS.md`
41+
42+### Reference
43+
44+See `WORKFLOWS.md` for:
45+- Architecture diagrams
46+- Workflow call graph
47+- Scheduled jobs & triggers
48+- Runners & secrets
4049
