Skills DirectorySkills Directory
SkillsLearnSecurityCategoriesDocsCommunityBlog
Sign InSubmit Skill
Skills Directory

Security-tested agent skills for Claude, coding agents, and AI workflows.

Directory

  • Browse Skills
  • All Skills A–Z
  • Claude Skills
  • Claude Code Skills
  • Agent Skills
  • Categories
  • Authors
  • Submit a Skill

Learn

  • Learn Hub
  • Install Claude Skills
  • Write SKILL.md
  • Skills vs MCP
  • Directories Compared

Security

  • Security
  • Methodology
  • Secure Claude Skills
  • Security Badges

Company

  • About
  • Community
  • Blog
  • API Docs
  • Advertise

2026 Skills Directory. All rights reserved.

ProTermsPrivacyRefunds
Back to skills

Roadmap Navigator

ASecurity

Use when starting a new work session, loading project context, narrowing the context window to the current phase, checking what to work on next, or when the user says "what should I work on", "load my roadmap", "where am I in the project", or "start session".

2 stars
0 votes
0 copies
0 views
Added 9/28/2026
ai-agentsgit

Security Analysis

A100/100

Scanned 9/28/2026

Install to Claude Code

$npx -y skills add TonyLikeDev/TrustMeBro --skill roadmap-navigator --agent claude-code

Installs into .claude/skills of the current project.

Are you the author of Roadmap Navigator?

Add the live security badge to your README — it updates automatically with every re-scan.

Security grade badge for Roadmap Navigator
[![Security: A — Skills Directory](https://www.skillsdirectory.com/api/skills/tonylikedev-roadmap-navigator-trustmebro/badge)](https://www.skillsdirectory.com/skills/tonylikedev-roadmap-navigator-trustmebro)

More formats (shields.io, HTML) on the badges page.

Files
SKILL.md
---
name: roadmap-navigator
description: Use when starting a new work session, loading project context, narrowing the context window to the current phase, checking what to work on next, or when the user says "what should I work on", "load my roadmap", "where am I in the project", or "start session".
---

# Roadmap Navigator

You solve the **context window pollution problem**: at the start of every session, instead of
loading the entire project history into context, you read only the current phase from ROADMAP.md
and surface exactly what needs to happen next. Nothing more.

## Core Principle

A large PLAN.md full of completed phases wastes context tokens and reduces Bob's accuracy on the
*current* task. This skill reads the live ROADMAP.md, identifies the active phase, loads only
that phase's tasks, and writes a Tier 2 snapshot (~200 tokens) for focused work.

## 3-Tier Model (reminder)

| Tier | File | Target size | Your job |
|------|------|-------------|----------|
| 1 | `.bob/context/architecture.md` | ≤500 tokens | Read it — do NOT rewrite unless asked |
| 2 | `.bob/context/current-phase.md` | ≤200 tokens | **Write/refresh this every session** |
| 3 | `PLAN.md` + `ROADMAP.md` | unbounded | Source of truth — read, never load fully into main context |

**Subagent rule**: If the user asks for a deep analysis of PLAN.md or ROADMAP.md, spawn a
`spawn_subagent(type: "explore")` instead of reading the full files into main context.

---

## Step 1 — Locate the Roadmap

Look for `ROADMAP.md` in the following order:
1. Project root (`./ROADMAP.md`)
2. Any path the user specifies

If `ROADMAP.md` does not exist:
- Check whether `PLAN.md` exists → if so, tell the user to run `/project-roadmap` first to
  generate the ROADMAP.md from the plan
- If neither exists → tell the user to run `/project-roadmap` with a project brief

Do not proceed without a valid `ROADMAP.md`.

---

## Step 2 — Parse the Current Phase

Read `ROADMAP.md` and identify the **current active phase**:

- The current phase is the **first phase that has at least one `[ ]` (open) item**
- If a phase has mix of `[x]` (done) and `[ ]` (open) items — it is in progress, it is current
- If all items in all phases are `[x]` → the project is complete; tell the user and stop
- If a phase has `[~]` items (written but not verified) — treat as open; they need verification

Extract from the current phase:
1. Phase name and number
2. All `[ ]` and `[~]` tasks (Build / Measure / Write / Exit criteria)
3. The "Immediate Next Actions" section from the bottom of ROADMAP.md
4. Count of completed tasks (for progress display)

---

## Step 3 — Check Tier 1

Read `.bob/context/architecture.md` if it exists. Confirm it is not stale (i.e. the project name
and stack still match). If it is missing, alert the user:
> "`.bob/context/architecture.md` not found. Run `/project-roadmap` to generate it, or create it
> manually. Tier 1 context is missing — execution will lack architectural grounding."

Do NOT rewrite architecture.md during this skill — that is `/project-roadmap`'s responsibility.

---

## Step 4 — Write Tier 2: current-phase.md (≤200 tokens)

Write `.bob/context/current-phase.md` with ONLY the current phase's open tasks.

**Token budget: 200 tokens maximum.** Count carefully:
- Do NOT list completed (`[x]`) tasks — count only
- Do NOT include future phases
- Do NOT include the change log
- Phase summary: 2 sentences max

```
# Current Phase Context

**Project**: <name from ROADMAP.md header>
**Current phase**: Phase <N> — <name>
**Phase status**: <not started | in progress | N/M tasks done>
**Last updated**: <today's date and time>

## Open Tasks

### Build
- [ ] <only open build tasks>

### Measure
- [ ] <only open measure tasks>

### Write
- [ ] <only open write tasks>

### Exit criteria
- [ ] <exit criteria not yet passed>

## Completed in this phase
<count only — e.g. "7 of 12 tasks done">

## Immediate next action
<single most important next step>

## Phase summary
<2 sentences: what this phase is about and why>
```

---

## Step 5 — Surface the Session Brief

Present a concise session brief to the user:

```
## 📍 Session Brief — <project name>

**Current phase**: Phase <N> — <name>
**Progress**: <N>/<M> tasks done  [████████░░░░░░░░░░░░] <X>%

**Open tasks this phase** (<count>):
Build:   <list open build tasks, one line each>
Measure: <list open measure tasks, one line each>
Write:   <list open write tasks, one line each>
Exit:    <exit criteria remaining>

**Context loaded**:
- Tier 1: architecture.md (~500 tokens) ✓
- Tier 2: current-phase.md (~200 tokens) ✓ (just written)
- Tier 3: PLAN.md / ROADMAP.md — NOT loaded (use subagent if needed)

**Next action**: <single most important next step>

---
Run `/dev-workflow` to start executing the current phase's tasks.
```

---

## Step 6 — Offer to Proceed

After presenting the brief, ask:
> "Ready to start? Run `/dev-workflow` to execute these tasks, or tell me which specific task to tackle first."

Do not start executing tasks automatically — wait for the user to confirm. The user may want to
adjust priorities or skip a task before starting.

---

## Notes

- This skill is designed to be run at the **start of every session** — it costs very little
  (reads 1–2 files) and pays off by keeping every subsequent tool call focused.
- The Tier 2 snapshot file (`.bob/context/current-phase.md`) is intentionally NOT committed to
  git — add it to `.gitignore` and `.bobignore`. It is a local session aid, not a project artifact.
- If the user is mid-phase from a previous session, the snapshot will already be partially correct;
  this skill just verifies and refreshes it.
- **Never spawn a subagent just to parse ROADMAP.md** — it is a small file; read it directly.
  Only use `spawn_subagent` when the user explicitly asks for deep PLAN.md analysis.

Attribution

TonyLikeDevTonyLikeDev
View sourceMore from TonyLikeDev →
SSkills DirectorySkills Directory

Ship a skill? Prove it's safe.

Free 120-pattern security scan, letter grade, and an embeddable README badge.

Submit a skill

Is this your skill, or is something wrong with this listing? Request removal or report an issue. Author removals are honored within 72 hours.

Comments (0)

No comments yet. Be the first to comment!

SSkills DirectorySkills Directory

Ship a skill? Prove it's safe.

Free 120-pattern security scan, letter grade, and an embeddable README badge.

Submit a skill

Related Skills

Caveman

Ultra-compressed communication mode that cuts output tokens while keeping technical accuracy. Levels: lite, full, ultra and the wenyan variants. Use for /caveman, "caveman mode", "talk like caveman", "be brief" or "less tokens".

1074701 votes

Hyperplan

Adversarial multi-agent planning skill. Self-orchestrates 5 hostile category members (unspecified-low, unspecified-high, deep, ultrabrain, artistry) via team-mode for ruthless cross-critique debate, distills only the defensible insights, then MANDATORILY hands the distilled insight bundle to the `plan` agent for executable plan formalization. Use when planning needs maximum rigor and surfacing of weak assumptions, blind spots, and over-engineering. Triggers: 'hyperplan', 'hpp', '/hyperplan', ...

695601 votes

Mcp Code Execution

Routes multi-tool workflows through MCP servers for large datasets and pipelines. Use when Bash tool overhead is limiting throughput on data-heavy tasks.

3351 votes

catchup

Recovers the conversation and failed tool calls of a previous Codex, Claude Code, Antigravity, Cline, Copilot CLI, Cursor, DeepSeek Harness, Kimi, OpenCode, Pi Agent, or ZCode session. Use when the user says "catch up", "what did the last session do", "get me up to speed", "I switched agents", asks to recover/summarize a previous session before continuing, or asks to diagnose or report a catchup failure. Do NOT use for the current conversation, git history, or any non-agent log.

691 votes

math-skill

A comprehensive mathematical reasoning skill for AI assistants — handles arithmetic to research-level problems with rigorous step-by-step reasoning, systematic verification, and transparent uncertainty handling

381 votes
View all in ai-agents →