AgenticSystemCore
Server Configuration
Describes the environment variables required to run the server.
| Name | Required | Description | Default |
|---|---|---|---|
No arguments | |||
Instructions
Guidance the server publishes about itself, which clients place ahead of the tool catalog so the model reads it before choosing anything.
This server publishes no instructions, or was last inspected before Glama recorded them.
Capabilities
Features and capabilities supported by this server
Protocol revision2025-11-25
| Capability | Details |
|---|---|
| tools | {} |
| prompts | {} |
| resources | {} |
| extensions | {
"com.agenticsystemcore/knowledge": {
"linkset": "http://localhost/.well-known/knowledge-linkset"
}
} |
Tools
Functions exposed to the LLM to take actions
| Name | Description |
|---|---|
| askA | Return the descriptions of the best-matching items for a question, citing their item IRIs; no model is called. |
| composeC | Run the AGSC-07 closure algebra over a selection of item slugs. |
| linksC | Return the typed Links authored on one item. |
| proposeC | Return the Proposal payload for one item — with task_state, a prepared board move (claim, progress, finish) as a patch; performs no write. |
| readC | Return one item by slug. |
| rememberC | Synthesize a conforming item — a new task on a board (kind task, cluster), a comment on one (about) — and return it as a Proposal payload. |
| searchC | Search this memory and return matching items. |
Prompts
Interactive templates invoked by user choice
| Name | Description |
|---|---|
| answer-from-this-memory-with-citations | answer from this memory with citations |
Resources
Contextual data attached and managed by the client
| Name | Description |
|---|---|
| example | |
| graph.jsonld | |
| llms.txt |
TDQS
Scored across 7 tools
Two pairs overlap meaningfully: ask vs search both retrieve items (differing mainly in whether a model/description output is used), and propose vs remember both return Proposal payloads (one retrieves an existing payload, one synthesizes a draft). read/links/compose are clearly distinct, and descriptions do clarify the boundaries, but an agent could still misselect within the retrieval and proposal pairs.
Nearly all tools are single lowercase verbs (ask, compose, propose, read, remember, search), giving a predictable style. The lone outlier is 'links', a noun that names a resource rather than an action, a minor deviation from an otherwise consistent convention.
Seven tools is a well-scoped set for a memory/knowledge core, covering retrieval, traversal, composition, and proposal drafting without redundancy. Each tool has a plausible distinct role in the workflow.
Query and draft coverage is good (read, search, ask, links, compose, remember, propose), but the surface stops at proposals: both propose and remember explicitly perform no write, and there is no apply/commit tool or any update/delete of existing items. Agents can draft but cannot persist changes, a notable lifecycle gap.