Roadmap logo

Roadmap

Community
davekilleen
roadmap

Review roadmap status, evidence freshness, blockers, and alignment

Overview

Publisherdavekilleen
RepositoryDex
Skill nameroadmap
Stars
481
Forks
130
Bundled files
Instructions only
Links
  • Markdown instructions

    A SKILL.md file the model loads on demand, so it only costs tokens when a request actually matches.

  • Works with any LLM

    AI skills are plain Markdown, not provider-specific code, so this works with GPT, Claude, Gemini, Grok, or a local model.

  • Self-contained

    Everything the model needs lives in the instructions — no extra files to sync.

  • Open source

    Published by davekilleen on GitHub. Read the source before you install it.

Installation

Install the Roadmap AI skill in TypingMind to use it with any LLM, or drop it into another agent that reads SKILL.md.

1

Install in TypingMind

TypingMind installs a skill straight from its GitHub folder — it reads SKILL.md, bundles the resource files, and stores the result locally.

  1. Open the app and go to Plugins → Skills.
  2. Choose "Install from GitHub".
  3. Paste the skill folder URL below and confirm.
  4. Enable the skill in any chat where you want it available.
Plugins → Skills → Add skill → From GitHub URL, then paste the folder URL and press Continue.
2

Install in another agent

Any agent that reads the Agent Skills format can use this skill — copy the folder into that agent's skills directory.

Claude Code — .claude/skills
git clone --depth 1 https://github.com/davekilleen/Dex.git /tmp/Dex
mkdir -p .claude/skills
cp -r /tmp/Dex/packages/dex-agent-plugin/skills/_available/product/roadmap .claude/skills/roadmap
Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Roadmap in any TypingMind chat and the model takes it from there. Its name and description sit in the system prompt, and the moment a request matches, the model loads the full instructions itself — you never invoke it by hand, and it costs no tokens until it is actually used.

The model loads Roadmap on its own as soon as a request matches it.

Works with any AI model

AI skills are plain Markdown instructions rather than provider-specific code, so Roadmap is not tied to the model it was written for. Install it once in TypingMind and use it with GPT-5, Claude, Gemini, Grok, DeepSeek, Mistral, Llama, or a local model you run yourself — all on your own API keys.

  • Loaded only when it is needed

    The system prompt carries just the name and description. The instructions are fetched on the first matching request, so an idle skill costs nothing.

  • Switch models mid-chat

    Because the skill is instructions rather than code, changing model does not break it — the next model reads the same SKILL.md.

Skill instructions

This is the SKILL.md content the model loads. Read it before installing — a skill is instructions your model will follow.

Purpose

Produce a dated, source-backed roadmap review across the confirmed project cohort. Surface explicit blockers, alignment gaps, evidence freshness, and unknown status without inventing an aggregate health judgment.

Usage

  • /roadmap — review the confirmed roadmap cohort
  • /roadmap [pillar] — filter by an exact configured pillar

Evidence, authority, and recovery

Treat the review as an evidence report, not a projection.

  • Use the canonical status source and status date from each project record. Record the path/ID, status date, and review as-of time. Filesystem modified time is only a discovery clue, never a replacement status date.
  • Cite source evidence for every status, blocker, milestone, pillar, alignment claim, and feedback item. If sources contradict, show both source dates and leave the field Unknown.
  • Unknown is distinct from blocked. Use Blocked only when the canonical source explicitly identifies an unresolved dependency preventing progress.
  • Define every denominator and disclose excluded projects. If the cohort or status is incomplete, report counts and coverage rather than a percentage or health score.
  • Never invent status, dates, blockers, dependencies, milestones, tags, feedback, counts, or health scores.
  • Keep analysis read-only. Preview requested changes, require explicit human authority, read back the target, and fail honestly on any mismatch.

Method

1. Confirm scope

Confirm:

  • review as-of date and timezone;
  • canonical project locations;
  • which project states belong on the roadmap;
  • requested pillar filter;
  • strategic pillar and quarterly-goal sources;
  • review window for stakeholder feedback;
  • configured freshness policy, if the user wants freshness labels.

Without a confirmed freshness policy, show the status age but label the assessment Unknown — no confirmed freshness policy.

2. Build the project ledger

For every discovered project capture:

FieldRequired evidence
Project identitystable path/ID
Statusexplicit value, status date, source
Milestoneexact dated source or Unknown
Blockerexplicit dependency and source, or Unknown
Pillarexact configured tag and source
Goal alignmentcited relationship, not name similarity
Feedbackdated speaker/source and quote-safe summary

Keep an Unchecked projects section for unreadable, duplicate, or contradictory records. Do not silently remove them from totals.

3. Classify without inference

  • In progress, Completed, and Blocked require an explicit canonical status.
  • An old status remains the last observed status plus a freshness caveat; age alone does not prove a blocker.
  • A keyword match is a lead. Read the surrounding source and distinguish an actual dependency from discussion or historical text.
  • Missing pillar means Unknown alignment, not automatically misaligned.
  • Meeting feedback corroborates a project record; it does not silently replace it.

Apply a freshness or alignment label only through a configured or cited rule. Record the rule's source and effective date.

4. Reconcile the roadmap

Verify:

  • discovered = checked + unchecked;
  • status counts reconcile to eligible checked projects;
  • pillar counts reconcile to projects with known valid pillars;
  • unknown-status and unknown-alignment projects are disclosed;
  • every blocker count points to explicit blocker evidence;
  • every percentage shows numerator, denominator, timeframe, and exclusions.

5. Recommend questions and actions

Separate:

  1. observed facts;
  2. contradictions and unknowns;
  3. policy-backed assessments;
  4. read-only recommendations;
  5. human decisions.

Do not infer priority from recency, meeting volume, or the number of mentions. If a priority source is missing, ask which source governs it.

Output contract

markdown
# Roadmap review

**As of:** [timestamp and timezone]
**Cohort:** [definition and source]
**Projects discovered / checked / unchecked:** [N / n / u]
**Freshness policy:** [source/date or Unknown]

## Active initiatives
### [Project]
- Canonical status: [value + source/status date or Unknown]
- Evidence freshness: [elapsed time; policy-backed label or Unknown]
- Next milestone: [value/source or Unknown]
- Pillar and goal alignment: [evidence or Unknown]
- Explicit blockers: [evidence or None observed; never inferred from silence]

## Attention and unknowns
- [Contradiction, missing status, explicit blocker, or unknown alignment]
- Evidence needed: [specific source or human decision]

## Stakeholder feedback
- [dated source and quote-safe summary; relationship to project]

## Evidence summary
Render one row per observed canonical status, including `Completed` when it is
observed. Do not force projects into a fixed list. Add `Unknown` for projects
without canonical status evidence, and reconcile every row to the checked cohort.

| Status | Count | Eligible denominator | Exclusions |
|---|---:|---:|---|
| [Observed canonical status, for example Completed] | [n] | [n/N] | [unknown/unchecked] |
| Unknown | [n] | [N] | [reasons] |

## Recommended actions
1. [Evidence-backed question or action; no write performed]

The output has no aggregate health score unless the user supplies a configured, source-backed rubric and explicitly asks for it.

Controlled changes

Before updating a project or creating a roadmap document:

  1. identify the authoritative target and current content;
  2. preview the exact operation and complete diff;
  3. name any linked project, goal, or pillar records that will remain unchanged;
  4. require explicit human confirmation;
  5. perform only the approved write;
  6. read back the target and compare it with the confirmed preview.

If writing fails or read-back differs, preserve prior content, report possible partial state, re-read the target, and present a corrected preview for fresh confirmation.

Frequently asked questions

What does the Roadmap AI skill do?

Review roadmap status, evidence freshness, blockers, and alignment

Why use Roadmap on TypingMind?

Because you install it once and use it with any model. Roadmap is plain Markdown rather than provider-specific code, so the same skill runs on GPT-5, Claude, Gemini, Grok, or a local model — and you can switch model mid-chat without it breaking. TypingMind runs on your own API keys, so you pay providers directly instead of a per-seat subscription, and your skills and chats stay in your own storage.

How do I install Roadmap in TypingMind?

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/davekilleen/Dex/tree/main/packages/dex-agent-plugin/skills/_available/product/roadmap. TypingMind reads its SKILL.md and installs it as a skill you can enable per chat.

Which AI models can use Roadmap?

Any model you connect in TypingMind. AI skills are plain Markdown instructions rather than provider-specific code, so GPT, Claude, Gemini, Grok, and local models can all load this skill when a request matches it.

How many AI models can I use with Roadmap?

As many as you like. As long as a model supports skills, you can use Roadmap with it — GPT, Claude, Gemini, Grok, DeepSeek, Mistral, Llama and more — all on TypingMind with your own API keys.

Is the Roadmap AI skill free?

It is published on GitHub by davekilleen. Check the repository for licensing terms. You only pay your own AI provider for the tokens you use.

What are AI skills?

An AI skill is a reusable instruction bundle that teaches an AI model how to do one specific task. It follows the open Agent Skills format: a SKILL.md file with a name and description, plus any scripts, templates or reference files the model may need. The model reads the instructions only when your request matches the skill, so an installed skill costs nothing until it is used.

How are AI skills different from plugins or MCP servers?

A plugin or MCP server gives a model new tools to call — code that runs somewhere and returns a result. An AI skill gives the model knowledge and process instead: how to approach a task, which steps to follow, what good output looks like. Skills are plain Markdown, so they need no server, no API key and no runtime, and they work with any model.

View all

Set up your own AI workspace now

Get notified about new features and future giveaways by subscribing to our newsletter 👇