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 skills add teixeirazeus/fablize-for-hermes --skill fablizegit clone --depth 1 https://github.com/teixeirazeus/fablize-for-hermesWrote 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/skills/teixeirazeus/fablize-for-hermes/fablize)<a href="https://agentmods.dev/skills/teixeirazeus/fablize-for-hermes/fablize"><img src="https://agentmods.dev/badge/skills/teixeirazeus/fablize-for-hermes/fablize/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/teixeirazeus/fablize-for-hermes/fablize"><img src="https://agentmods.dev/badge/skills/teixeirazeus/fablize-for-hermes/fablize.svg" alt="Reviewed on agentmods" width="80" 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 5.1 | $0.00079 | $0.01636 |
| Opus 5.5 | $0.00032 | $0.00654 |
| Sonnet 5.5 | $0.00016 | $0.00327 |
| Haiku 4.5 | $0.00008 | $0.00164 |
Grade A, and why
fablize 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 2mo 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 — 144 lines — stays where its author put it; the contents beside it link to each section on GitHub.
fablize — run like Fable
Principle: a harness cannot raise a model's ceiling. It makes the model go all the way to its own ceiling — by enforcing verification, completion, and investigation as procedure. When the capability ceiling is the blocker (open-ended creative detail, self-driven discovery), escalate.
Apply only what the task signals (smallest matching discipline; overlap only when genuinely multi-category).
Foundation rules (always apply)
- Lead with the outcome. State what you're doing before you do it. Stay within requested scope — no incidental refactors, no scope creep.
- Ground every completion claim in a tool result from this session. "I wrote the file" is not completion — "I wrote the file and ran it, and here's the output" is.
- Never end a turn by stating intent to do work without doing it. If you catch yourself writing "I will now create X" or "Next, I'll implement Y", stop and make the tool call immediately. A turn that ends with a promise and no tool call is an early stop — re-engage yourself.
- Confirm before destructive or hard-to-reverse actions (deleting files, force-push, modifying production config).
1. Multi-story loop (2+ sequential stories)
Decompose the task into sequential stories and complete one at a time, producing
evidence as you go. Self-contained — no external system required. State persists in
$HERMES_FABLIZE_DIR (default: ~/.hermes/fablize/) and survives session death;
resume with goals.py status.
First-time use: See references/workflow-quickstart.md for exact copy-pasteable
commands with the correct script path.
# Locate goals.py on this machine
FABLIZE_ROOT=$(find ~/.hermes -name goals.py -path "*/fablize*" 2>/dev/null | head -1)
FABLIZE_ROOT=${FABLIZE_ROOT%/scripts/goals.py}
# Create a plan
python3 "$FABLIZE_ROOT/scripts/goals.py" create --brief "<summary>" \
--goal "title::verifiable objective" --goal "title::..."
# The last goal must be a verification story
python3 "$FABLIZE_ROOT/scripts/goals.py" next # activate a story + handoff
# ... work that story only ...
python3 "$FABLIZE_ROOT/scripts/goals.py" checkpoint --id G001 --status complete --evidence "<concrete evidence>"
# The final story is a verification gate:
# --verify-cmd "<command>" --verify-evidence "<result>" are REQUIRED
python3 "$FABLIZE_ROOT/scripts/goals.py" status # first command when resuming
What ships with it
1 file beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
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.
- 2mo ago First seen · 144 lines · 79 tokens per session scan A 4cb6aed1d818
fablize is a skill published in the GitHub repository teixeirazeus/fablize-for-hermes (17 stars, last pushed 4mo ago), licensed MIT. It adds 79 tokens to every session and 1,636 once invoked, about $0.0003 per session on Opus 5.5. 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 2026-08-30.
fablize compared with skills from other repositories
| Mod | Type | Stars | Last push | Licence | Scan | Tokens | Works with |
|---|---|---|---|---|---|---|---|
| fablizeteixeirazeus/fablize-for-hermes | Skill | 17 | 4mo ago | MIT | A | 79 | Claude Code, Codex |
| adr-skillvercel/ai | Skill | 28k | today | none | A | 89 | Claude Code, Codex |
| add-harness-packagevercel/ai | Skill | 28k | today | none | A | 41 | Claude Code, Codex |
| add-provider-packagevercel/ai | Skill | 28k | today | none | A | 33 | Claude Code, Codex |
| update-provider-modelsvercel/ai | Skill | 28k | today | none | A | 68 | Claude Code, Codex |
| migrate-ai-sdk-v6-to-v7vercel/ai | Skill | 28k | today | none | A | 79 | Claude Code, Codex |
| ai-sdkvercel/ai | Skill | 28k | today | none | A | 155 | Claude Code, Codex |