opencode-skills-collection 3.0.46 → 3.0.48
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
- package/bundled-skills/.antigravity-install-manifest.json +15 -1
- package/bundled-skills/2slides-ppt-generator/SKILL.md +1 -1
- package/bundled-skills/2slides-ppt-generator/scripts/create_pdf_slides.py +2 -1
- package/bundled-skills/2slides-ppt-generator/scripts/generate_narration.py +2 -1
- package/bundled-skills/2slides-ppt-generator/scripts/generate_slides.py +13 -7
- package/bundled-skills/accint-solve/SKILL.md +205 -0
- package/bundled-skills/android-cli/SKILL.md +239 -0
- package/bundled-skills/android-cli/references/interact.md +83 -0
- package/bundled-skills/android-cli/references/journeys.md +105 -0
- package/bundled-skills/android-dev/references/hybrid.md +7 -4
- package/bundled-skills/android-dev/references/react-native.md +5 -2
- package/bundled-skills/atlas-contract/SKILL.md +4 -4
- package/bundled-skills/atlas-ledger/SKILL.md +10 -7
- package/bundled-skills/bun-development/SKILL.md +1 -1
- package/bundled-skills/cloud-penetration-testing/SKILL.md +1 -1
- package/bundled-skills/codebase-to-wordpress-converter/SKILL.md +1 -0
- package/bundled-skills/codex-fable5/SKILL.md +154 -0
- package/bundled-skills/docs/integrations/jetski-cortex.md +3 -3
- package/bundled-skills/docs/integrations/jetski-gemini-loader/README.md +1 -1
- package/bundled-skills/docs/maintainers/repo-growth-seo.md +3 -3
- package/bundled-skills/docs/maintainers/skills-update-guide.md +1 -1
- package/bundled-skills/docs/users/bundles.md +1 -1
- package/bundled-skills/docs/users/claude-code-skills.md +1 -1
- package/bundled-skills/docs/users/gemini-cli-skills.md +1 -1
- package/bundled-skills/docs/users/getting-started.md +1 -1
- package/bundled-skills/docs/users/kiro-integration.md +1 -1
- package/bundled-skills/docs/users/usage.md +4 -4
- package/bundled-skills/docs/users/visual-guide.md +4 -4
- package/bundled-skills/dos-verify-done-claims/SKILL.md +173 -0
- package/bundled-skills/ecl-harness-engineer/LICENSE +21 -0
- package/bundled-skills/ecl-harness-engineer/SKILL.md +714 -0
- package/bundled-skills/ecl-harness-engineer/agents/analyzer.md +119 -0
- package/bundled-skills/ecl-harness-engineer/agents/auditor.md +212 -0
- package/bundled-skills/ecl-harness-engineer/agents/creator-config.md +343 -0
- package/bundled-skills/ecl-harness-engineer/agents/creator-docs.md +201 -0
- package/bundled-skills/ecl-harness-engineer/agents/creator-linters.md +123 -0
- package/bundled-skills/ecl-harness-engineer/references/adapters/adapter-schema.md +204 -0
- package/bundled-skills/ecl-harness-engineer/references/adapters/generic.md +156 -0
- package/bundled-skills/ecl-harness-engineer/references/adapters/go.md +212 -0
- package/bundled-skills/ecl-harness-engineer/references/adapters/java.md +205 -0
- package/bundled-skills/ecl-harness-engineer/references/adapters/python.md +225 -0
- package/bundled-skills/ecl-harness-engineer/references/adapters/rust.md +220 -0
- package/bundled-skills/ecl-harness-engineer/references/adapters/typescript.md +245 -0
- package/bundled-skills/ecl-harness-engineer/references/architecture-diagrams.md +420 -0
- package/bundled-skills/ecl-harness-engineer/references/audit-templates.md +649 -0
- package/bundled-skills/ecl-harness-engineer/references/capability-registry.md +485 -0
- package/bundled-skills/ecl-harness-engineer/references/darwin-eval-prompts.md +373 -0
- package/bundled-skills/ecl-harness-engineer/references/documentation-templates.md +741 -0
- package/bundled-skills/ecl-harness-engineer/references/durability-patterns.md +423 -0
- package/bundled-skills/ecl-harness-engineer/references/ecl-harness.md +1431 -0
- package/bundled-skills/ecl-harness-engineer/references/environment-config-guide.md +534 -0
- package/bundled-skills/ecl-harness-engineer/references/environment-detection-guide.md +751 -0
- package/bundled-skills/ecl-harness-engineer/references/eval-templates.md +377 -0
- package/bundled-skills/ecl-harness-engineer/references/gc-templates.md +798 -0
- package/bundled-skills/ecl-harness-engineer/references/greenfield-templates.md +1385 -0
- package/bundled-skills/ecl-harness-engineer/references/linter-templates.md +448 -0
- package/bundled-skills/ecl-harness-engineer/references/observability-templates.md +315 -0
- package/bundled-skills/efficient-web-research/SKILL.md +320 -0
- package/bundled-skills/environment-setup-guide/SKILL.md +2 -2
- package/bundled-skills/evolution/SKILL.md +1 -1
- package/bundled-skills/gitops-workflow/SKILL.md +1 -1
- package/bundled-skills/linkerd-patterns/SKILL.md +1 -1
- package/bundled-skills/loki-mode/examples/todo-app-generated/frontend/package-lock.json +504 -1317
- package/bundled-skills/loki-mode/examples/todo-app-generated/frontend/package.json +2 -2
- package/bundled-skills/lovable-cleanup/SKILL.md +416 -0
- package/bundled-skills/monopoly/SKILL.md +397 -0
- package/bundled-skills/monopoly/patterns/SKILL.md +331 -0
- package/bundled-skills/monopoly/scale-benchmarks/SKILL.md +174 -0
- package/bundled-skills/monopoly/security-checklist/SKILL.md +69 -0
- package/bundled-skills/monopoly/tech-matrix/SKILL.md +268 -0
- package/bundled-skills/pagespeed-enhancer/SKILL.md +579 -0
- package/bundled-skills/polis-protocol/SKILL.md +6 -3
- package/bundled-skills/sharp-coder/SKILL.md +131 -0
- package/bundled-skills/unship/SKILL.md +11 -5
- package/bundled-skills/uv-package-manager/resources/implementation-playbook.md +1 -1
- package/bundled-skills/varlock/SKILL.md +2 -2
- package/package.json +1 -1
- package/skills_index.json +314 -4
|
@@ -34,7 +34,7 @@ antigravity-awesome-skills/
|
|
|
34
34
|
├── 📄 CONTRIBUTING.md ← Contributor workflow
|
|
35
35
|
├── 📄 CATALOG.md ← Full generated catalog
|
|
36
36
|
│
|
|
37
|
-
├── 📁 skills/ ← 1,
|
|
37
|
+
├── 📁 skills/ ← 1,555+ skills live here
|
|
38
38
|
│ │
|
|
39
39
|
│ ├── 📁 brainstorming/
|
|
40
40
|
│ │ └── 📄 SKILL.md ← Skill definition
|
|
@@ -47,7 +47,7 @@ antigravity-awesome-skills/
|
|
|
47
47
|
│ │ └── 📁 2d-games/
|
|
48
48
|
│ │ └── 📄 SKILL.md ← Nested skills also supported
|
|
49
49
|
│ │
|
|
50
|
-
│ └── ... (1,
|
|
50
|
+
│ └── ... (1,555+ total)
|
|
51
51
|
│
|
|
52
52
|
├── 📁 apps/
|
|
53
53
|
│ └── 📁 web-app/ ← Interactive browser
|
|
@@ -100,7 +100,7 @@ antigravity-awesome-skills/
|
|
|
100
100
|
|
|
101
101
|
```
|
|
102
102
|
┌─────────────────────────┐
|
|
103
|
-
│ 1,
|
|
103
|
+
│ 1,555+ SKILLS │
|
|
104
104
|
└────────────┬────────────┘
|
|
105
105
|
│
|
|
106
106
|
┌────────────────────────┼────────────────────────┐
|
|
@@ -201,7 +201,7 @@ If you want a workspace-style manual install instead, cloning into `.agent/skill
|
|
|
201
201
|
│ ├── 📁 brainstorming/ │
|
|
202
202
|
│ ├── 📁 stripe-integration/ │
|
|
203
203
|
│ ├── 📁 react-best-practices/ │
|
|
204
|
-
│ └── ... (1,
|
|
204
|
+
│ └── ... (1,555+ total) │
|
|
205
205
|
└─────────────────────────────────────────┘
|
|
206
206
|
```
|
|
207
207
|
|
|
@@ -0,0 +1,173 @@
|
|
|
1
|
+
---
|
|
2
|
+
name: dos-verify-done-claims
|
|
3
|
+
description: "Before accepting an agent's 'done / shipped / fixed' claim, verify it against ground truth (git ancestry + the commit's own diff) using the DOS kernel's `dos verify` and `dos commit-audit` — never the agent's own narration."
|
|
4
|
+
category: quality
|
|
5
|
+
risk: safe
|
|
6
|
+
source: community
|
|
7
|
+
source_repo: anthony-chaudhary/dos-kernel
|
|
8
|
+
source_type: community
|
|
9
|
+
date_added: "2026-06-12"
|
|
10
|
+
author: anthony-chaudhary
|
|
11
|
+
tags: [verification, git, ai-agents, trust, quality-gate]
|
|
12
|
+
tools: [claude, cursor, gemini]
|
|
13
|
+
license: "MIT"
|
|
14
|
+
license_source: "https://github.com/anthony-chaudhary/dos-kernel/blob/master/LICENSE"
|
|
15
|
+
---
|
|
16
|
+
|
|
17
|
+
# Verify done-claims against ground truth, not the agent's word
|
|
18
|
+
|
|
19
|
+
## Overview
|
|
20
|
+
|
|
21
|
+
When an AI agent says "done", "shipped", or "fixed", that is a **claim**, not a
|
|
22
|
+
fact — and a claim the agent checks by re-reading its own work is *consistency,
|
|
23
|
+
not grounding*. This skill replaces that self-report with a verdict from a
|
|
24
|
+
witness the agent did not author: it shells the **DOS kernel** (`dos verify`,
|
|
25
|
+
`dos commit-audit`) to confirm the claimed effect from git ancestry and the
|
|
26
|
+
commit's actual diff. DOS is deterministic — no API key, no LLM. The verdict is
|
|
27
|
+
git-only and offline as used here; the one exception is `dos verify` in a
|
|
28
|
+
workspace that wires a CI oracle, which `--no-ci` suppresses (see Security &
|
|
29
|
+
Safety Notes).
|
|
30
|
+
|
|
31
|
+
This skill adapts the DOS reference "witness-claim" pattern
|
|
32
|
+
(`anthony-chaudhary/dos-kernel`) into a host-agnostic screenplay.
|
|
33
|
+
|
|
34
|
+
## When to Use This Skill
|
|
35
|
+
|
|
36
|
+
- Use when an agent reports a task/phase/feature as **complete** and you want
|
|
37
|
+
that "done" confirmed from evidence before building on it.
|
|
38
|
+
- Use right after a commit, to confirm the commit's **message matches its diff**
|
|
39
|
+
(catch a `fix:` that only touched a README, or a "tests pass" that deleted the
|
|
40
|
+
assertions).
|
|
41
|
+
- Use when folding many sub-agents' results — verify each claimed effect instead
|
|
42
|
+
of trusting the return string.
|
|
43
|
+
- **Do not** use it to judge whether code is *correct* — that is what the test
|
|
44
|
+
suite is for. This skill checks did-the-claimed-thing-actually-ship.
|
|
45
|
+
|
|
46
|
+
## How It Works
|
|
47
|
+
|
|
48
|
+
### Step 1: Install the kernel (once)
|
|
49
|
+
|
|
50
|
+
```bash
|
|
51
|
+
pip install dos-kernel # provides the `dos` CLI; deterministic, no key
|
|
52
|
+
```
|
|
53
|
+
|
|
54
|
+
### Step 2: Audit the latest commit's claim vs its diff
|
|
55
|
+
|
|
56
|
+
A commit subject is forgeable (whoever wrote the message authored it); the files
|
|
57
|
+
it touched are not (git did). `dos commit-audit` grades the subject against the
|
|
58
|
+
actual diff:
|
|
59
|
+
|
|
60
|
+
```bash
|
|
61
|
+
dos commit-audit --workspace . HEAD --json
|
|
62
|
+
```
|
|
63
|
+
|
|
64
|
+
`commit-audit --json` prints a JSON **array** of audited commits (one element
|
|
65
|
+
even for a single `HEAD`), so read `verdict` from the first element — e.g.
|
|
66
|
+
`dos commit-audit --workspace . HEAD --json | jq -r '.[0].verdict'`. (Without
|
|
67
|
+
`--json` the same verdict prints as a one-line text row: `· OK …`,
|
|
68
|
+
`⚑ UNWITNESSED …`, or `· abstain …`.) The verdicts are: `OK` (the diff backs the
|
|
69
|
+
claim's *kind*), `CLAIM_UNWITNESSED` (the subject's claim is not evidenced by the
|
|
70
|
+
diff — treat the "done" as unproven), or `ABSTAIN`. This judges the *kind* of
|
|
71
|
+
change, never correctness — run the tests for that.
|
|
72
|
+
|
|
73
|
+
### Step 3: Verify a named phase actually shipped
|
|
74
|
+
|
|
75
|
+
If the agent claims a specific plan/phase landed, confirm it from git history
|
|
76
|
+
rather than the transcript:
|
|
77
|
+
|
|
78
|
+
```bash
|
|
79
|
+
dos verify --workspace . PLAN PHASE --json --no-ci
|
|
80
|
+
```
|
|
81
|
+
|
|
82
|
+
`--no-ci` keeps the verdict git-only (see the Security note below). With `--json`
|
|
83
|
+
you get the `shipped` and `source` fields. (The default text form prints
|
|
84
|
+
`SHIPPED PLAN PHASE (via grep)` or `NOT_SHIPPED PLAN PHASE (via none)` — the same
|
|
85
|
+
verdict, and the process exit code is non-zero when not shipped.)
|
|
86
|
+
|
|
87
|
+
Grade `shipped: true` by the `source`, because git fallback grades itself by
|
|
88
|
+
**forgeability** — and forgeable evidence is exactly what this skill exists to
|
|
89
|
+
distrust:
|
|
90
|
+
|
|
91
|
+
- `registry` or `grep-artifact` — **non-forgeable** (a registry row, or an
|
|
92
|
+
artefact/diff rung). This closes the claim.
|
|
93
|
+
- `grep-subject` (or bare `grep`) — **forgeable**: a commit *subject* or body
|
|
94
|
+
carried the phase token, which an agent can write without doing the work (even
|
|
95
|
+
on an empty commit). Treat this as *shipped-per-the-subject*, not confirmed —
|
|
96
|
+
corroborate it (run `dos commit-audit` on that commit, below) before you close.
|
|
97
|
+
- `none` — no positive evidence; accept as "not shipped", not as a tool failure.
|
|
98
|
+
|
|
99
|
+
### Step 4: Fold only confirmed effects
|
|
100
|
+
|
|
101
|
+
Accept the agent's "done" **only** when Step 2/3 corroborate it. If
|
|
102
|
+
`CLAIM_UNWITNESSED` or `shipped: false`, the work is not done regardless of how
|
|
103
|
+
confidently the agent narrated it — send it back.
|
|
104
|
+
|
|
105
|
+
## Examples
|
|
106
|
+
|
|
107
|
+
### Example 1: gate an agent's "I fixed the bug" claim
|
|
108
|
+
|
|
109
|
+
```bash
|
|
110
|
+
# The agent committed and said it's fixed. Check the diff backs the claim.
|
|
111
|
+
# commit-audit --json returns an array, so read the first element's verdict:
|
|
112
|
+
dos commit-audit --workspace . HEAD --json | jq -r '.[0].verdict'
|
|
113
|
+
# OK -> the change is of the claimed kind; now run the tests
|
|
114
|
+
# CLAIM_UNWITNESSED -> the commit doesn't do what it says; reject
|
|
115
|
+
```
|
|
116
|
+
|
|
117
|
+
### Example 2: confirm a feature phase shipped before closing a ticket
|
|
118
|
+
|
|
119
|
+
```bash
|
|
120
|
+
dos verify --workspace . AUTH AUTH2 --json --no-ci
|
|
121
|
+
# shipped: true, source: registry|grep-artifact -> non-forgeable; safe to close
|
|
122
|
+
# shipped: true, source: grep-subject|grep -> forgeable subject/body match;
|
|
123
|
+
# shipped-per-the-subject only -> corroborate with commit-audit before closing
|
|
124
|
+
# shipped: false, source: none -> no evidence; keep the ticket open
|
|
125
|
+
```
|
|
126
|
+
|
|
127
|
+
## Best Practices
|
|
128
|
+
|
|
129
|
+
- ✅ Run `dos commit-audit HEAD` immediately after every agent commit.
|
|
130
|
+
- ✅ Treat `source: none` / `CLAIM_UNWITNESSED` as "not done", not as a tool error.
|
|
131
|
+
- ✅ Close a claim on a **non-forgeable** `source` (`registry`, `grep-artifact`).
|
|
132
|
+
Treat `grep-subject` / bare `grep` as forgeable (an agent can write the subject
|
|
133
|
+
text) — corroborate before closing.
|
|
134
|
+
- ✅ Keep the test suite as the separate correctness gate — this skill checks shipping, not correctness.
|
|
135
|
+
- ❌ Don't accept a "done" because the agent's prose was confident.
|
|
136
|
+
- ❌ Don't use this to replace code review or testing.
|
|
137
|
+
|
|
138
|
+
## Limitations
|
|
139
|
+
|
|
140
|
+
- This skill does not replace environment-specific validation, testing, or expert review.
|
|
141
|
+
- It checks whether a claimed change *shipped* / matches its diff — not whether the code is *correct*.
|
|
142
|
+
- `dos verify` reads git history; in a repo with no commits there is nothing to witness (it will honestly report `source: none`).
|
|
143
|
+
- Stop and ask for clarification if required inputs (a git repo, the `dos` CLI) are missing.
|
|
144
|
+
|
|
145
|
+
## Security & Safety Notes
|
|
146
|
+
|
|
147
|
+
- This skill runs shell commands: `pip install dos-kernel` and the read-only
|
|
148
|
+
`dos` verbs (`dos commit-audit`, `dos verify`). These verbs never **mutate**
|
|
149
|
+
the repo or push. `dos commit-audit` only reads git history and the working
|
|
150
|
+
tree (no network). `dos verify` is also git-only **unless** the workspace has
|
|
151
|
+
wired a CI oracle (`[verify] non_git_oracle` in its `dos.toml`), in which case
|
|
152
|
+
it may shell a network check (e.g. `gh api`) for the verdict — pass `--no-ci`
|
|
153
|
+
(as the examples above do) to force the git-only path and guarantee no network.
|
|
154
|
+
- `pip install dos-kernel` installs from PyPI. The distribution name is
|
|
155
|
+
`dos-kernel` (the bare `dos` on PyPI is an unrelated package — do not install
|
|
156
|
+
it). Pin a version in locked environments.
|
|
157
|
+
- Run in the repository you intend to adjudicate; the `--workspace .` argument
|
|
158
|
+
scopes every verdict to that repo.
|
|
159
|
+
|
|
160
|
+
## Common Pitfalls
|
|
161
|
+
|
|
162
|
+
- **Problem:** `dos verify` returns `source: none` and it looks like a failure.
|
|
163
|
+
**Solution:** That is the honest "no evidence" verdict — it means the phase has
|
|
164
|
+
no ship commit, so the claim is unproven. Re-stamp the real commit or keep the
|
|
165
|
+
task open.
|
|
166
|
+
- **Problem:** Installing the wrong package.
|
|
167
|
+
**Solution:** The PyPI name is `dos-kernel`, not `dos`.
|
|
168
|
+
|
|
169
|
+
## Related Skills
|
|
170
|
+
|
|
171
|
+
- The upstream DOS reference screenplays (`dos-witness-claim`, `dos-goal-gate`)
|
|
172
|
+
in `anthony-chaudhary/dos-kernel` cover the multi-agent fan-out and
|
|
173
|
+
self-stopping-agent variants of this same witness discipline.
|
|
@@ -0,0 +1,21 @@
|
|
|
1
|
+
MIT License
|
|
2
|
+
|
|
3
|
+
Copyright (c) 2026 ecl-harness-engineer contributors
|
|
4
|
+
|
|
5
|
+
Permission is hereby granted, free of charge, to any person obtaining a copy
|
|
6
|
+
of this software and associated documentation files (the "Software"), to deal
|
|
7
|
+
in the Software without restriction, including without limitation the rights
|
|
8
|
+
to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
|
|
9
|
+
copies of the Software, and to permit persons to whom the Software is
|
|
10
|
+
furnished to do so, subject to the following conditions:
|
|
11
|
+
|
|
12
|
+
The above copyright notice and this permission notice shall be included in all
|
|
13
|
+
copies or substantial portions of the Software.
|
|
14
|
+
|
|
15
|
+
THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
|
|
16
|
+
IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
|
|
17
|
+
FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
|
|
18
|
+
AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
|
|
19
|
+
LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
|
|
20
|
+
OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
|
|
21
|
+
SOFTWARE.
|