Thanks to visit codestin.com
Credit goes to github.com

Skip to content

New1Direction/SKTPG

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

3 Commits
 
 
 
 
 
 
 
 

Repository files navigation

SKTPG — Skate Where The Puck Is Going

A directional intelligence skill for AI agents.

Skate where the puck is going, but show your evidence.

Most people ask: What is happening?

/sktpg asks: What is this becoming?

Better: What does this force next?

Best: What becomes obvious 6–18 months from now that is still non-obvious today?


What it does

/sktpg forces you to stop thinking about "what is happening now?" and start thinking in:

  • Weak signals — early motion before consensus forms
  • Hype vs real motion — narrative separation with evidence
  • Direction of travel — vector, not point
  • Incentive maps — who is pulled, who is squeezed, who is forced to move
  • Bottleneck maps — current bottleneck → next bottleneck → who profits
  • Second-order questions — 10× cheaper? regulated? default infra?
  • 6–18 month forecasts — base / bull / bear / wild card, all falsifiable
  • Action maps — what to do in 24h / 7d / 30d / before consensus
  • Tracking signals — green / yellow / red scoreboard
  • The SKTPG Thesis — compact 6-line verdict

Every conclusion is tagged by evidence strength: Confirmed · Likely · Speculative · Contradicted · Unknown


Install

SKTPG works on any platform that supports AI agent skills — Hermes, Claude Code, Cursor, Copilot, Codex, etc.

Hermes Agent

hermes skill install https://github.com/New1Direction/SKTPG

Claude Code

Clone the repo, then link the skill:

git clone https://github.com/New1Direction/SKTPG ~/.claude/skills/sktpg

Or use the plugin manifest if you publish via marketplace.

Any platform (manual)

Clone the repo and symlink SKILL.md into your agent's skills directory:

git clone https://github.com/New1Direction/SKTPG ~/sktpg
# then symlink SKILL.md or the whole folder into your agent's skills path

Then use:

/sktpg What's the real signal behind the AI agent framework explosion?
/sktpg https://github.com/some-project
/sktpg API inference prices just dropped 80%

Golden tests

The skill ships with verification cases in references/:

Test Input Expected score
Hype-only "This AI agent will replace all workers." 0–20 (noise)
Real weak signal "Small teams replacing QA with browser agents, humans for final approval" 41–70 (watchlist)
Bottleneck shift "API provider cuts inference prices by 80%" 60–85 (actionable)
Repo weak signal Repo with prompts/mocks, no runtime proof Identify scaffold vs tool
Protocol incentive L2 changes fee distribution favoring stakers Map forced moves

Scoring system

Category Points Question
Signal strength 20 Real weak signals beyond hype?
Incentive alignment 20 Are actors forced to move?
Bottleneck shift 15 Real constraint being removed?
Second-order upside 15 Unlocks new markets/workflows?
Timing advantage 15 Action now matter before consensus?
Falsifiability 10 Thesis trackable and killable?
Personal edge 5 User has plausible advantage?

0–20 Noise · 21–40 Interesting · 41–60 Watchlist · 61–80 Actionable · 81–100 Urgent opportunity


Guardrails

  • Do not overpredict — directional theses with falsification criteria
  • Do not summarize only — explain what it forces next
  • Do not confuse attention with motion — usage > hype, adoption > funding, workflow > demo
  • Do not ignore bottlenecks — current bottleneck AND next bottleneck
  • Do not skip action — always answer "what should I do before this becomes obvious?"

License

MIT

About

Directional intelligence for AI agents. Skate where the puck is going — weak-signal detection, hype separation, incentive mapping, and 6-18 month forecasts.

Topics

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors