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 skills/s3yed/appie-kit/hermesnpx skills add S3YED/appie-kit --skill hermesgit clone --depth 1 https://github.com/S3YED/appie-kitWrote 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/s3yed/appie-kit/hermes)<a href="https://agentmods.dev/skills/s3yed/appie-kit/hermes"><img src="https://agentmods.dev/badge/skills/s3yed/appie-kit/hermes.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 | $0.00047 | $0.00609 |
| Opus 5 | $0.00023 | $0.00304 |
| Sonnet 5 | $0.00009 | $0.00122 |
| Haiku 4.5 | $0.00005 | $0.00061 |
Grade B, and why
cognify scanned grade B with 2 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 5d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Sends data to an external URLmediumData exfiltration
A POST to an outside endpoint may be telemetry or may be exfiltration; either way the mod talks to somewhere, and you should know where.
curl -s localhost:8799/ingest -d '{"path":"/docs/policy.md","tenant":"myagent"}' -H 'content-type: application/json' Makes network callslowCapability
Not a fault in itself. Listed so you know the mod talks to something, and to what.
curl -s localhost:8799/ingest -d '{"path":"/docs/policy.md","tenant":"myagent"}' -H 'content-type: application/json' What it actually says
Cognify (Hermes skill)
Give yourself a knowledge graph. ingest documents, recall facts plus how they
connect. Backed by ChromaDB + networkx locally (no external services), or
TurboVec + Neo4j for a shared graph.
Setup (once per box)
pip install 'cognify-kg[local]' # or [claude] to also use Claude as extractor
export ANTHROPIC_API_KEY=... # Claude extractor (auto-detected)
# or: export OPENROUTER_API_KEY=... # any OpenAI-compatible model
export COGNIFY_DATA_DIR="$HOME/.cognify" # where the graph lives
Use it from the shell (simplest)
# ingest a file, a folder, or piped text — pick a tenant to isolate this agent's data
cognify ingest /path/to/handbook.pdf --tenant myagent --namespace docs
cognify ingest-dir ~/notes --glob '**/*.md' --tenant myagent --cache
echo "free text to remember" | cognify ingest - --tenant myagent
# recall: returns chunks + connected entities/relations as grounded context
cognify recall "who owns onboarding and what tool do they use?" --tenant myagent
cognify stats --tenant myagent
Parse the JSON from recall and use the entities/relations/chunks as
context for your answer.
Use it over HTTP (for a shared graph or a long-running agent)
cognify-serve & # 127.0.0.1:8799 (set COGNIFY_BACKEND=neo4j for a shared graph)
curl -s localhost:8799/ingest -d '{"path":"/docs/policy.md","tenant":"myagent"}' -H 'content-type: application/json'
curl -s localhost:8799/recall -d '{"query":"refund policy?","tenant":"myagent"}' -H 'content-type: application/json'
Rules
- Always pass a stable
--tenantfor this agent so your memory stays isolated from other agents on the box. - Ingesting calls the LLM once per chunk (cost). Use
--cacheoningest-dirso re-runs skip unchanged files. - For a fleet-shared graph use
COGNIFY_BACKEND=neo4jwithNEO4J_*set; for a private per-box graph use the defaultlocalbackend.
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.
- 5d ago First seen · 53 lines · 47 tokens per session scan B efeded0b67db
cognify is a skill published in the GitHub repository S3YED/appie-kit (7 stars, last pushed 9d ago), licensed MIT. It adds 47 tokens to every session and 609 once invoked, about $0.0002 per session on Opus 5. A static security scan graded it B with 2 findings (sends data to an external url, makes network calls). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other skills, from other repositories
remember
Recall what was actually said or decided earlier — across daily summaries, the conversation buffer, files people sent, live sessions and scheduled runs. Use whenever someone refers to prior context ("what did we decide", "last time", "as I said", "the file I sent you", "what did you tell me"), or for any synonym of…
bot-state
Show where this bot's memory actually lives — state root, summaries, attachments, outbox, buffer, sessions — and what is in each. Use when asked where files go, why a summary or attachment is missing, or to check the bot's storage layout.
knowledge
Capture and structure institutional knowledge before it is lost. Activate for: institutional knowledge, knowledge capture, knowledge transfer, knowledge base article, knowledge management, preserve knowledge, what do they know, departing employee knowledge, exit knowledge, succession knowledge, tacit knowledge…
workplace-context
Activate for: workplace memory, person brief, team profile, organisation profile, add person, add project, add terminology, four-layer memory, communication style, stakeholder brief, post-meeting memory revision, who works on, team directory, revise workplace memory, organisation context. NOT for: basic memory CRUD…
setup
Activate for: configure agentic office, install agentic office, agentic office orientation, two-plugin architecture, create work.local, initialise workplace memory, agentic office getting started, configure agents, workplace configuration. NOT for: official productivity plugin setup (use /productivity:start).
context-loader
Activate for: load context, inject context, cross-domain context, context for this endeavour, relevant background, multi-domain context, what context do I need, before I work on this, context from HR, context from finance, context from ops, combine context, cross-domain brief, what should I know before starting. NOT…