Usage
Which Claude Code features this work actually leans on — mined from session transcripts. Metadata only: tool names and timing, never commands or content.
By category
9 kindsActivity
peak 4341/dayBy project
3 shownRecommended, but rarely used
High-leverage features the playbook flags that this work barely touches.
- code-review — Whenever a skill matches the task; it beats hand-assembling the steps.
- update-config — Whenever a skill matches the task; it beats hand-assembling the steps.
- frontend-design — Whenever a skill matches the task; it beats hand-assembling the steps.
- schedule — Whenever a skill matches the task; it beats hand-assembling the steps.
- brainstorm — Whenever a skill matches the task; it beats hand-assembling the steps.
- persona-design — Whenever a skill matches the task; it beats hand-assembling the steps.
- Workflow — Audits, migrations, multi-file reviews, or any fan-out where you want structure and an adversarial verify pass rather than one serial pass.
- Subagent — Independent work you can delegate and keep only the conclusion of — broad searches, or parallel tasks that would otherwise blow up the main context.
Routines you could automate
10 foundTool-sequences this work repeats by hand, mined from session history. Each scaffold is a starting point to adapt, not a finished workflow.
command:/clear→·codeAutomation scaffold
A recurring sequence: 1. run /clear 2. edit files / run code & tests Worth saving as a slash-command if you find yourself repeating it.
·code→Skill:pre-push-review→·codeAutomation scaffold
You already run the `pre-push-review` skill inside this sequence: 1. edit files / run code & tests 2. run the pre-push-review skill 3. edit files / run code & tests Invoke it directly with `/pre-push-review`, and consider a saved slash-command that runs the surrounding steps so the whole sequence is one action.
·code→Agent:general-purpose→·codeAutomation scaffold
This is an edit → verify loop you repeat by hand:
1. edit files / run code & tests
2. spawn a general-purpose subagent
3. edit files / run code & tests
Automate it with `/loop` (re-run the check on an interval) or a small Workflow that re-runs the verification until it passes:
```js
// /loop <your verify command> — simplest option
// or a Workflow:
while (!passed) { await agent('run the verification, report pass/fail', {schema: VERDICT}); }
```
_Starting scaffold — adapt before running._Skill:pre-push-review→·code→Agent:general-purpose→·codeAutomation scaffold
You already run the `pre-push-review` skill inside this sequence: 1. run the pre-push-review skill 2. edit files / run code & tests 3. spawn a general-purpose subagent 4. edit files / run code & tests Invoke it directly with `/pre-push-review`, and consider a saved slash-command that runs the surrounding steps so the whole sequence is one action.
·code→Skill:pre-push-review→·code→Agent:general-purposeAutomation scaffold
You already run the `pre-push-review` skill inside this sequence: 1. edit files / run code & tests 2. run the pre-push-review skill 3. edit files / run code & tests 4. spawn a general-purpose subagent Invoke it directly with `/pre-push-review`, and consider a saved slash-command that runs the surrounding steps so the whole sequence is one action.
Task→·code→Task→·codeAutomation scaffold
This is an edit → verify loop you repeat by hand:
1. run a background task
2. edit files / run code & tests
3. run a background task
4. edit files / run code & tests
Automate it with `/loop` (re-run the check on an interval) or a small Workflow that re-runs the verification until it passes:
```js
// /loop <your verify command> — simplest option
// or a Workflow:
while (!passed) { await agent('run the verification, report pass/fail', {schema: VERDICT}); }
```
_Starting scaffold — adapt before running._·code→Task→·code→TaskAutomation scaffold
This is an edit → verify loop you repeat by hand:
1. edit files / run code & tests
2. run a background task
3. edit files / run code & tests
4. run a background task
Automate it with `/loop` (re-run the check on an interval) or a small Workflow that re-runs the verification until it passes:
```js
// /loop <your verify command> — simplest option
// or a Workflow:
while (!passed) { await agent('run the verification, report pass/fail', {schema: VERDICT}); }
```
_Starting scaffold — adapt before running._·code→mcp:playwright→·codeAutomation scaffold
This is an edit → verify loop you repeat by hand:
1. edit files / run code & tests
2. drive playwright over MCP
3. edit files / run code & tests
Automate it with `/loop` (re-run the check on an interval) or a small Workflow that re-runs the verification until it passes:
```js
// /loop <your verify command> — simplest option
// or a Workflow:
while (!passed) { await agent('run the verification, report pass/fail', {schema: VERDICT}); }
```
_Starting scaffold — adapt before running._·code→Task→·code→Skill:pre-push-reviewAutomation scaffold
You already run the `pre-push-review` skill inside this sequence: 1. edit files / run code & tests 2. run a background task 3. edit files / run code & tests 4. run the pre-push-review skill Invoke it directly with `/pre-push-review`, and consider a saved slash-command that runs the surrounding steps so the whole sequence is one action.
Task→·code→Skill:pre-push-review→·codeAutomation scaffold
You already run the `pre-push-review` skill inside this sequence: 1. run a background task 2. edit files / run code & tests 3. run the pre-push-review skill 4. edit files / run code & tests Invoke it directly with `/pre-push-review`, and consider a saved slash-command that runs the surrounding steps so the whole sequence is one action.
Features
54 usedRuns a shell command in the working directory.
Exact string-replacement edit on a previously read file.
Reads a file (text, image, PDF, or notebook) from disk.
Creates a new file or fully replaces an existing one.
Creates and tracks an asynchronous task that runs across turns.
Calls an external system over MCP (browser automation, Drive, Gmail, Notion, etc.).
Loads deferred tool schemas on demand so they become callable.
Presents a structured multiple-choice question to the user.
Calls an external system over MCP (browser automation, Drive, Gmail, Notion, etc.).
Calls an external system over MCP (browser automation, Drive, Gmail, Notion, etc.).
Invokes a packaged skill / slash-command — a reusable procedure with its own instructions.
A slash-command typed into the prompt.
Calls an external system over MCP (browser automation, Drive, Gmail, Notion, etc.).
Calls an external system over MCP (browser automation, Drive, Gmail, Notion, etc.).
Calls an external system over MCP (browser automation, Drive, Gmail, Notion, etc.).
Calls an external system over MCP (browser automation, Drive, Gmail, Notion, etc.).
Calls an external system over MCP (browser automation, Drive, Gmail, Notion, etc.).
Fetches and reads a specific URL.
Calls an external system over MCP (browser automation, Drive, Gmail, Notion, etc.).
Calls an external system over MCP (browser automation, Drive, Gmail, Notion, etc.).
Searches the web.
Invokes a packaged skill / slash-command — a reusable procedure with its own instructions.
Calls an external system over MCP (browser automation, Drive, Gmail, Notion, etc.).
A slash-command typed into the prompt.
Invokes a packaged skill / slash-command — a reusable procedure with its own instructions.
A slash-command typed into the prompt.
Calls an external system over MCP (browser automation, Drive, Gmail, Notion, etc.).
A slash-command typed into the prompt.
Invokes a packaged skill / slash-command — a reusable procedure with its own instructions.
Calls an external system over MCP (browser automation, Drive, Gmail, Notion, etc.).
A slash-command typed into the prompt.
Invokes a packaged skill / slash-command — a reusable procedure with its own instructions.
Invokes a packaged skill / slash-command — a reusable procedure with its own instructions.
Calls an external system over MCP (browser automation, Drive, Gmail, Notion, etc.).
Invokes a packaged skill / slash-command — a reusable procedure with its own instructions.
Invokes a packaged skill / slash-command — a reusable procedure with its own instructions.
Calls an external system over MCP (browser automation, Drive, Gmail, Notion, etc.).
Calls an external system over MCP (browser automation, Drive, Gmail, Notion, etc.).
A slash-command typed into the prompt.
Invokes a packaged skill / slash-command — a reusable procedure with its own instructions.
Invokes a packaged skill / slash-command — a reusable procedure with its own instructions.