prompt-vars
Use prompt vars for the active plan and design files, plus merged project/global agent-mode vars.
Skill by judepayne
Chalk Labs
Need help deploying prompt-vars?
Chalk Labs sets up agents like this inside real teams: connected to your tools, tested on your data, with a human in the loop where it matters.
Install
Works with Claude Code, Claude apps, Codex, Cursor, Gemini CLI, GitHub Copilot, VS Code, OpenCode, Goose, Windsurf. Read the repository before you give any tool access to your data.
npx agentmarket install skill/judepayne-prompt-vars
Asks which of your tools to add it to (Claude Code, Codex, Cursor, VS Code, Windsurf, Gemini CLI, Claude Desktop) and shows the change first. See https://agentmart.dev/cli.
Get the weekly diff
The weekly diff: one email a week with new agent tools, notable updates and grade changes, plus security alerts for tools you follow. Double opt-in, one-click unsubscribe.
About
Use cases: design, writing, agent-building
Reviews
Write a reviewNo reviews yet. Installed it? Reviews come from verified installers only. How reviews work
Permissions
What this skill can do and what it asks for, from its SKILL.md. Checked Oct 25, 2026.
- Tools
- No tools found
- Secrets
- None asked for
- Install scripts
- None found
Network hosts it talks to 1
codeload.github.com· found in install command
Files it touches
~/.claude/skills/prompt-vars
What's inside
- Skill name
prompt-vars- Description
- Use prompt vars for the active plan and design files, plus merged project/global agent-mode vars. Use when writing or updating agent prompts that should adapt to plan/design state, or when you need to inspect or mutate those vars at runtime.
- License
- MIT
- Instructions
- 189 lines in SKILL.md · view on GitHub
Files (1)
- SKILL.md
SKILL.md (first lines)
# Prompt Vars
This environment provides:
- built-in derived prompt vars for the active plan and design files
- project vars from `.pi/agent-mode-vars.json`
- global fallback vars from `~/.pi/agent/agent-mode-vars.json`
- a project-local write-location config in `.pi/agent-mode-vars-config.json`
- a `vars` tool and `/vars` command for runtime inspection, bootstrap, and mutation
If you need the full wiring and implementation details, read:
- `extensions/z-prompt-vars/README.md`
## What they are for
Use prompt vars to make prompts adapt to the workspace and current runtime state without hardcoding absolute paths.
Typical uses:
- agent mode definitions
- other system-prompt-like text artifacts
- runtime inspection when you need the live value of a var
- workspace-local or user-global configuration that should persist across sessions
## Built-in derived prompt vars
These are computed at runtime and are read-only:
- `${plan.path}`
- `${plan.exists}`
- `${plan.active}`
- `${design.path}`
- `${design.exists}`
- `${design.active}`
- `${plan}`
- `${design}`
Meaning:
- `*.path` is the canonical active file path
- `*.exists` is `true` or `false`
- `*.active` is `true` when the corresponding mode is currently active (`planner` for plan, `designer` for design)
- `${plan}` and `${design}` expand to short status sentences
The source path vars for those built-ins are:
- `paths.plan`
- `paths.design`
If `paths.plan` or `paths.design` is unset, the extension falls back to:
- `.pi/plans/active.md`
- `.pi/designs/active.md`
## Read and write behavior
Read precedence:
1. project `.pi/<vars-file-name>`
2. global `~/.pi/agent/<vars-file-name>`
Project vars override global vars.
Writes go to the location selected by:
- `.pi/agent-mode-vars-config.json`
The config contains:
```json
{
"pi-location": "project",
"vars-file-name": "agent-mode-vars.json"
}
```
`pi-location` may be:
- `project`
- `global`
`vars-file-name` is optional and overrides the default vars filename while keeping the project/global directories fixed.
If a path is mistakenly provided, the extension keeps only the final path segment.
## Workspace vars
Examples of useful stored keys:
- `project.name`
- `feature.flag`
- `release.owner`
- `paths.design`
- `subagents.dispatch.defaultContext`
- `automode.enabled`
- `gate.auto.*` for pi-gate local auto-approval configuration
Recommended default:
- `subagents.dispatch.defaultContext = "fresh"`
- `automode.enabled = false`
- `gate.auto.enabled = false`
- `gate.auto.startOnSession = false`
Use `fork` only as an exception when a subagent truly needs prior session context.
## Runtime interfaces
### The `vars` tool
Agents should use the `vars` tool when they need a live value or need to update vars programmatically.
Examples:
- `vars({ action: "bootstrap" })`
- `vars({ action: "list" })`
- `vars({ action: "get", key: "plan.path" })`
- `vars({ action: "get", key: "project.name" })`
- `vars({ action: "set", key: "project.name", value: "Prompt Vars" })`
- `vars({ action: "set", key: "flags", value: { beta: true, rollout: 25 } })`
- `vars({ action: "set", key: "subagents.dispatch.defaultContext", value: "fresh" })`
- `vars({ action: "set", key: "subagents.dispatch.defaultContext", value: "fork" })`
- `vars({ action: "unset", key: "project.name" })`
- `vars({ action: "location" })`
- `vars({ action: "location", value: "global" })`
### The `/vars` command
Users can do the same interactively.
Examples:
- `/vars bootstrap`From judepayne/picode, MIT.
More from judepayne/picode
- 0.587score
orchestrate-subagents
judepayne
Use mediated subagent delegation through delegate_subagent and delegate_subagent_status.
SkillAFreshness: A★ 1Claude Code +9 - 0.494score
planning-workflow
judepayne
Use a structured planning workflow to produce an implementation-ready Builder handoff grounded in the actual repository.
SkillCFreshness: C★ 1Claude Code +9 - 0.491score
automode-designer
judepayne
Designer-mode automode handoff rules for explicit user-started Design → Plan → Build automation.
SkillCFreshness: C★ 1Claude Code +9 - 0.491score
automode-planner
judepayne
Planner-mode automode rules for turning the active design into a Builder-ready plan and handing off to Builder.
SkillCFreshness: C★ 1Claude Code +9 - 0.491score
partner-reviewer
judepayne
Run the bounded Partner Reviewer workflow for implementations that changed five or more unique files.
SkillCFreshness: C★ 1Claude Code +9 - 0.486score
automode-builder
judepayne
Builder-mode automode rules for implementing the active plan and clearing automode on stop.
SkillCFreshness: C★ 1Claude Code +9 - 0.432score
pr-management
judepayne
Use the team-lead subagent to run a staged PR workflow.
SkillDFreshness: D★ 1Claude Code +9 - 0.426score
karpathy-coding-discipline
judepayne
Behavioral guidelines to reduce common LLM coding mistakes.
SkillDFreshness: D★ 1Claude Code +9
Signals
- Stars
- 1
- Forks
- 0
- Open issues
- 0
- Last commit (this folder)
- Jul 10, 2026 (3 months ago)
From GitHub (fetched Oct 24, 2026).
Details
- License
- MIT
- Repository
- github.com/judepayne/picode
- First seen
- Oct 24, 2026
- Last checked
- Oct 24, 2026
- Source
- GitHubgithub-graphql
More from judepayne
- 0.587score
orchestrate-subagents
judepayne
Use mediated subagent delegation through delegate_subagent and delegate_subagent_status.
SkillAFreshness: A★ 1Claude Code +9 - 0.494score
planning-workflow
judepayne
Use a structured planning workflow to produce an implementation-ready Builder handoff grounded in the actual repository.
SkillCFreshness: C★ 1Claude Code +9 - 0.491score
automode-designer
judepayne
Designer-mode automode handoff rules for explicit user-started Design → Plan → Build automation.
SkillCFreshness: C★ 1Claude Code +9 - 0.491score
automode-planner
judepayne
Planner-mode automode rules for turning the active design into a Builder-ready plan and handing off to Builder.
SkillCFreshness: C★ 1Claude Code +9 - 0.491score
partner-reviewer
judepayne
Run the bounded Partner Reviewer workflow for implementations that changed five or more unique files.
SkillCFreshness: C★ 1Claude Code +9 - 0.486score
automode-builder
judepayne
Builder-mode automode rules for implementing the active plan and clearing automode on stop.
SkillCFreshness: C★ 1Claude Code +9
Done for you
Rather not set it up yourself?
Chalk Labs scopes, builds and runs agents for your team: the task audit, the eval harness, the integrations and the weekly ops report. Fixed scope, priced before we start.
See how we build AI agents →Read the case studies
Chalk Labs is a two-founder AI and growth studio. We build AI agents, CRMs and MVPs for Web3 and AI companies, and have shipped 50+ product and token launches.
- launches shipped
- 50+
- launches shipped
- media placements
- 500+
- media placements
- founders on every account
- 2
- founders on every account
Press we've placed: Forbes · CoinDesk · Decrypt · Cointelegraph · The Block
Rahil Jain, Co-founder, Chalk Labs. Ex-Goldman Sachs engineer. You talk to the person who builds it.
Get prompt-vars running in your business
Tell us what you want to automate. A founder replies within one working day with a plan, not a pitch deck.
Cite this page
prompt-vars skill. Agent Mart. https://agentmart.dev/skills/judepayne-prompt-vars (last updated 2026-10-24)