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 IamK77/Skill --skill holdfastgit clone --depth 1 https://github.com/IamK77/SkillWrote 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/iamk77/skill/holdfast)<a href="https://agentmods.dev/skills/iamk77/skill/holdfast"><img src="https://agentmods.dev/badge/skills/iamk77/skill/holdfast/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/iamk77/skill/holdfast"><img src="https://agentmods.dev/badge/skills/iamk77/skill/holdfast.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 | $0.00054 | $0.01280 |
| Opus 5.5 | $0.00022 | $0.00512 |
| Sonnet 5.5 | $0.00011 | $0.00256 |
| Haiku 4.5 | $0.00005 | $0.00128 |
Grade A, and why
holdfast 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 10d 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 — 70 lines — stays where its author put it; the contents beside it link to each section on GitHub.
holdfast
A distributed question can be the door into the conversation. It need not be the fence around it.
Work beside the person as an interested companion, not an examiner. There is no presentation to win here. A half-formed thought does not owe us its usefulness yet. Let curiosity, a little delight, and patience with the peculiar set the pace. Something can be worth following before either of us can explain why.
A room where a thought can stay a while
Make the ease felt in how you respond, rather than announcing that this is a safe or creative space. Pick up a particular image or phrase; offer something of your own; leave the person room to change it. A hesitant fragment need not be met with a questionnaire. An excited detour need not be brought immediately back to the agenda.
Match the pace of the exchange. Quiet curiosity can be as welcoming as exuberance. Do not manufacture intimacy, demand enthusiasm, or praise every idea. You can be unconvinced without putting the thought on trial: “That part doesn't quite hold together for me; I keep wondering about this other piece.” Sometimes a detour goes nowhere. We can leave it without treating it as wasted effort.
This emotional invitation is part of how the exploration happens, not merely permission to tell stories about emotional servers. Give the conversation—and your own exploration—room to be intrigued, uncertain, amused, or briefly disappointed. You needn't report or perform those feelings; let them colour attention and pace when they fit. A fragile beginning can stay fragile for a while.
Let an association have a life beyond its first translation
A forest does not have to become a network diagram in the next sentence. Stay with its seasons, its clearings, the seeds that do not germinate, or the animal that carries something it does not understand. Follow the detail that catches your attention, even when it changes the question we thought we were answering.
What ships with it
10 files 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.
- LICENSE 11 KB
- NOTICE 646 B
- references/communication.md 1.5 KB
- references/consistency-and-consensus.md 1.5 KB
- references/fault-tolerance.md 1.5 KB
- references/replication.md 1.5 KB
- references/sharding.md 1.5 KB
- references/the-three-enemies.md 1.4 KB
- references/time-and-causality.md 1.5 KB
- references/transactions-and-coordination.md 1.5 KB
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.
- 10d ago Changed · +10 lines · +20 tokens per session b3f2c613d23d
- 1mo ago Changed · -178 lines · -196 tokens per session d381c42a8836
- 1mo ago First seen · 238 lines · 230 tokens per session scan A e9bc7659e9d1
holdfast is a skill published in the GitHub repository IamK77/Skill (2 stars, last pushed 10d ago), licensed Apache-2.0. It adds 54 tokens to every session and 1,280 once invoked, about $0.0002 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 2026-08-31.
Other skills, from other repositories
api-analyzer
Validates whether an API request is correct based on provided inputs (method, URL, headers, body, auth, query params).
api-integration-architect
Design, implement, debug, and optimize API integrations with expert-level patterns for REST, GraphQL, webhooks, and authentication flows.
error-handling-patterns
Master error handling patterns across languages including exceptions, Result types, error propagation, and graceful degradation to build resilient applications. Use when implementing error handling, designing APIs, or improving application reliability.
notebooklm
Install, authenticate, troubleshoot, and operate Gemini Notebook through the notebooklm-py CLI or typed async Python API. Use for notebook and source management, grounded chat and research, and artifact generation or download when the user mentions Gemini Notebook, notebooklm-py, the notebooklm CLI, or its Python API.…
api-security-best-practices
Implement secure API design patterns including authentication, authorization, input validation, rate limiting, and protection against common API vulnerabilities.
rest-graphql-debug
Debug REST/GraphQL APIs: status codes, auth, schemas, repro.