Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx agentmods add instructions/epistates/sparx/claude-mdgit clone --depth 1 https://github.com/Epistates/sparXWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/instructions/epistates/sparx/claude-md)<a href="https://agentmods.dev/instructions/epistates/sparx/claude-md"><img src="https://agentmods.dev/badge/instructions/epistates/sparx/claude-md.svg" alt="Measured on agentmods" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5 | $0.01345 | $0.01345 |
| Opus 5 | $0.00673 | $0.00673 |
| Sonnet 5 | $0.00269 | $0.00269 |
| Haiku 4.5 | $0.00135 | $0.00135 |
Grade A, and why
sparX CLAUDE.md scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 4d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 114 lines — stays where its author put it; the contents beside it link to each section on GitHub.
sparX — X Post Optimization System
Turn Claude Code into the ultimate X (Twitter) content engine. Every skill in this project is grounded in how the Phoenix algorithm actually scores posts.
How This Project Works
This is a skill-based system. You interact with it through /slash-commands:
| Skill | Purpose |
|---|---|
/compose |
Draft algorithm-optimized single posts |
/thread |
Create high-engagement thread sequences |
/optimize |
Score and rewrite existing drafts for max reach |
/hooks |
Generate scroll-stopping opening lines |
/schedule |
Get optimal posting times based on data |
/research |
Research trending topics in your niche |
/review |
Analyze post-publish performance |
/strategy |
Plan weekly content strategy |
/repurpose |
Transform content across formats (post→thread, thread→post, etc.) |
/media |
Plan and create visual content (screenshots, GIFs, Remotion video) |
/post |
Publish posts/threads directly to X via chrome browser automation |
/analyze |
Read real engagement analytics from X via browser |
/engage |
Manage replies and conversations on X via browser (75–150× multiplier) |
/setup |
Install optional integrations (Remotion, OpenTweet, last30days, X API) |
/test |
Verify sparX setup is working correctly |
Phoenix Algorithm — Essential Context
Always apply this knowledge when working with X content in this project.
The Scoring Formula
Every post shown in "For You" is scored by the Phoenix transformer:
Final Score = Σ (weight × P(action)) - negative_penalties + diversity_adjustment
Phoenix predicts 15-18 engagement probabilities per user×post pair. The weights are not equal — this hierarchy determines reach:
| Signal | Weight vs Like | Why It Matters |
|---|---|---|
| Author replies to comments | 75–150× | Creates conversation velocity → massive distribution |
| Deep replies / threaded convos | 13–27× | Strongest standard positive signal |
| Repost / Quote | ~20× | High trust signal from other users |
| Bookmark | ~10× | "Silent like" — extremely predictive of quality |
| Dwell time (time spent reading) | Very high | Core transformer signal; longer = much better |
| Video completion % | High | Media gets natural boost |
| Like | 1× | Baseline — least valuable positive signal |
| Block/Mute/Report/Not interested | Heavy penalty | Asymmetric — one bad interaction tanks future reach |
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 4d ago First seen · 114 lines · 1,345 tokens per session scan A 046e7387205c
sparX CLAUDE.md is an instructions file published in the GitHub repository Epistates/sparX (3 stars, last pushed 5mo ago), licensed MIT. It adds 1,345 tokens to every session, about $0.0067 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other instructions, from other repositories
Skill_Seekers CLAUDE.md
Instructions for yusufkaraaslan/Skill_Seekers, covering claude.md, project overview, essential commands, required before running tests or cli (src/ layout) and run all tests (never skip - all must pass before commits).
stealth-browser-mcp AGENTS.md
Instructions for vibheksoni/stealth-browser-mcp, covering agent instructions, start here, recommended proxy services, environment and validation.
octocode CLAUDE.md
Claude Code instructions for bgauryy/octocode: Read AGENTS.md. Discover repository skills under .agents/skills and activate only those matching the task.
che-ical-mcp CLAUDE.md
Claude Code instructions for PsychQuant/che-ical-mcp, covering spectra instructions, use /spectra: skills when, workflow, parked changes and mcp 功能驗證的表面選擇（.claude/rules/mcp-binary-version-testing.md）.
ai-workflow AGENTS.md
Instructions for cunhaax/ai-workflow, covering ai workflow template, rules — non-negotiable, project overview, commands and architecture.
substack-mcp-plus CLAUDE.md
Instructions for ty13r/substack-mcp-plus, covering claude.md, quick command reference (use these!), always activate virtual environment first, run tests (use python3!) and code quality.