Skills DirectorySkills Directory
SkillsLearnSecurityCategoriesDocsBlogPro
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
  • Chrome Extension
  • Skill Manager

Company

  • About
  • Community
  • Blog
  • API Docs
  • Advertise

2026 Skills Directory. All rights reserved.

ProTermsPrivacyRefunds
Back to skills

S Waydriver

ASecurity

Reusable control plane for discovering, mapping, coordinating, and resuming a foggy or context-spanning effort. Use when the route itself needs durable decisions, dependencies, frontier state, or Wayfinder-compatible continuation; callers supply domain-specific delivery and verification policy.

5 stars
0 votes
0 copies
4 views
Added 9/19/2026
ai-agentssecuritydocumentation

Security Analysis

A100/100

Pro scans all 7 files and shows the line behind each finding

Scanned 10/6/2026

$npx -y skills add simonasrazm/skills --skill s-waydriver --agent claude-code

Installs into .claude/skills of the current project.

Are you the author of S Waydriver?

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

Security grade badge for S Waydriver
[![Security: A — Skills Directory](https://www.skillsdirectory.com/api/skills/simonasrazm-s-waydriver/badge)](https://www.skillsdirectory.com/skills/simonasrazm-s-waydriver)

More formats (shields.io, HTML) on the badges page. Keep it an A: scan every change in CI with Pro.

Download with Pro
Files
SKILL.md
---
name: s-waydriver
description: Reusable control plane for discovering, mapping, coordinating, and resuming a foggy or context-spanning effort. Use when the route itself needs durable decisions, dependencies, frontier state, or Wayfinder-compatible continuation; callers supply domain-specific delivery and verification policy.
---

# S Waydriver

A destination has arrived but the route is not fully visible. Chart only what can be seen, resolve the decisions that clear the fog, then drive acceptance-bearing work until the destination is reached or no safe route remains.

## Preserve the decision map

The decision map is a canonical index, not a store. A decision lives in exactly one ticket or lasting decision record; the map gists and links it. Refer to maps and units by their human-readable names, with ids carried inside links.

Keep the upstream-compatible decision surface:

```markdown
## Destination

<the accepted end state>

## Notes

<domain, standing preferences, compatibility base, and capabilities to consult>

## Decisions so far

- [<resolved decision name>](link): <one-line consequence>

## Not yet specified

<in-scope fog not yet sharp enough to ask as a precise question>

## Out of scope

<conscious boundaries held outside the frontier>
```

Use the configured issue tracker. If none exists, use `.scratch/<effort>/map.md` with decision tickets under `.scratch/<effort>/issues/`. Read [tracker operations](references/tracker.md) when creating or adapting the carrier.

## Clear fog proportionally

When work introduces or changes domain concepts, language, relationships or rules, read and apply [domain modeling](../s-domain-modeling/SKILL.md).

A decision ticket is a precise question whose answer changes the safe route. It is not an implementation slice. Preserve Wayfinder's four types:

- **Grilling:** preference, intent, or domain choices. Apply [autonomous discovery](references/discovery.md); involve the person only when the consequential answer remains theirs.
- **Research:** facts outside the working context. Use `smart-shot` when available, or the same primary-source, multi-domain evidence outcome with available tools.
- **Prototype:** a concrete experiment can answer what discussion cannot. Apply [prototype guidance](references/prototype.md).
- **Task:** action or access needed before a decision can be made. Complete it when authorized; otherwise persist and communicate the blocker while independent work continues.

Fog becomes a ticket when its question can be stated precisely. Resolving a ticket graduates newly visible questions and removes that patch from **Not yet specified**. Out-of-scope work is closed as a boundary, not carried as fog.

On continuation, reconcile material user prompts, feedback and discovery results with recorded commitments. Preserve stated requirements and reasons with source pointers; attribute each choice to its human or agent author. Keep agent defaults and inferred rationale distinct from human approval. Apply explicit corrections in place, update affected acceptance, and reopen unresolved conflicts. Update current requirements in place when an accepted correction replaces them. Preserve decision reasons rather than transcripts of deliberation. Persist material changes in the map, affected units and lasting decisions before responding, including decision-only responses. These records must agree with the latest material input.

Classify lasting records as **product** (behavior, feature, content, UX, business rule) or **architecture** (technical structure, platform, data, integration, security boundary, scalability, operations). Follow the project's decision-record location; otherwise place them under `docs/decisions/product/` or `docs/decisions/architecture/`. Both human and autonomous lasting decisions use these records. Include the choice, decision maker, rationale, supporting evidence, remaining assumptions and the originating unit or surgical-task slug. Link changed documentation and decision records back from that unit. When a decision changes, update its existing record at its stable path with the current choice, reason, decision maker and evidence. Align affected documentation, specifications, glossary and acceptance with that choice; distinguish accepted direction from implementation still pending. Before replacing a decision, verify that its prior accepted state is recoverable in version history; record any history gap before proceeding. Keep review evidence bound to the contract and candidate it actually assessed. An unresolved conflict stays open with its affected work identified. Reconcile accepted discovery results into existing project documentation; the ticket and map link the authoritative record.

## Drive delivery separately

When enough decisions are clear to build, create or resume a linked delivery map; keep implementation units separate from Wayfinder decision tickets. Locally use `.scratch/<effort>/delivery.md`. In an issue tracker, use a linked delivery-map issue with its own acceptance-bearing child units.

Delivery units are coherent vertical changes with observable acceptance, real dependencies, candidate identity, and current evidence. The frontier is the open, unblocked, unclaimed set. Keep one maintained-product mutation unit active unless isolation and merge/revalidation are demonstrated. Non-mutating research and checks may overlap when their inputs are frozen and their outputs retrievable.

A caller may supply implementation and verification capabilities. Waydriver records their units, candidate identities, evidence, and blockers without imposing a domain-specific delivery or verification policy. A repair creates a new candidate and reopens affected evidence.

When starting or resuming a delivery unit, read [execution runs](references/execution-runs.md) and bind its execution record to the unit and accepted contract. Keep pending inputs, repairs and acceptance resumable through that binding.

## Continue safely

Each context begins from the destination, frontier, active candidate, decisions, fog, blockers, and evidence, not replayed conversation. Claim before mutating and persist resolution before taking another unit.

One mutation-bearing unit at a time is the focus invariant, not a forced session ending. A session may continue into the next safe frontier unit while context remains reliable. At a context boundary, leave enough state for a fresh executor to continue without guessing.

Drive until the destination has current acceptance evidence, every remaining route is blocked, or a consequential preference or authority decision cannot be derived safely. Ask a person only for that irreducible input; resume independent frontier work before and after the request when possible.

Attribution

simonasrazmsimonasrazm
View sourceSee grades on GitHubMore from simonasrazm →
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

Terse caveman voice: answer first, fluff gone, every technical fact kept. Use for /caveman, "caveman mode", "talk like caveman", "be brief", "less tokens". Stays on until "stop caveman" or "normal mode".

1100021 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', ...

698621 votes

Writing Skills

Create and manage Claude Code skills in HASH repository following Anthropic best practices. Use when creating new skills, modifying skill-rules.json, understanding trigger patterns, working with hooks, debugging skill activation, or implementing progressive disclosure. Covers skill structure, YAML frontmatter, trigger types (keywords, intent patterns), UserPromptSubmit hook, and the 500-line rule. Includes validation and debugging with SKILL_DEBUG. Examples include rust-error-stack, cargo-dep...

3931 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.

3421 votes

catchup

Recovers the conversation and failed tool calls of a previous Codex, Amp, Claude Code, Antigravity, Cline, Copilot CLI, Cursor, DeepSeek Harness, Grok Build, 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.

741 votes
View all in ai-agents →