profile-project
Server Configuration
Describes the environment variables required to run the server.
| Name | Required | Description | Default |
|---|---|---|---|
| PROFILE_PROJECT_CHROMADB__PATH | No | Path for ChromaDB storage | .profile_project/chroma |
| PROFILE_PROJECT_OPENAI_API_KEY | No | OpenAI API key (secret) | |
| PROFILE_PROJECT_PINECONE_API_KEY | No | Pinecone API key (secret) | |
| PROFILE_PROJECT_VECTORSTORE__BACKEND | No | Vectorstore backend (unset prompts for selection) | |
| PROFILE_PROJECT_DEFAULT_EMBEDDINGS_METHOD | No | Default embeddings method (unset prompts for selection) |
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 | {
"listChanged": false
} |
| prompts | {
"listChanged": false
} |
| resources | {
"subscribe": false,
"listChanged": false
} |
| experimental | {} |
Tools
Functions exposed to the LLM to take actions
| Name | Description |
|---|---|
| pp_config_pathD | – |
| pp_config_showD | – |
| pp_config_getD | – |
| pp_config_setD | – |
| pp_config_validateD | – |
| pp_init_projectD | – |
| pp_init_runD | – |
| pp_next_phasesD | – |
| pp_start_phaseD | – |
| pp_complete_phaseD | – |
| pp_fail_phaseD | – |
| pp_retry_phaseD | – |
| pp_run_statusD | – |
| pp_list_runsD | – |
| pp_discover_sourcesD | – |
| pp_list_sourcesD | – |
| pp_get_sourceD | – |
| pp_add_sourceD | – |
| pp_store_artifactD | – |
| pp_load_artifactD | – |
| pp_list_artifactsD | – |
| pp_validate_artifactD | – |
| pp_index_buildD | Chunk -> embed -> upsert agent pages into the store (§9.3/§11). |
| pp_index_rebuildC | Full geometry-safe re-index (reset + rebuild, §10.4/§11). |
| pp_queryB | Embed query -> vector search -> ranked, attributed hits (§11). |
| pp_index_statusC | Store stats. In-handler read-only init guard (§11/§6b.3). Count-only by design: a status read reuses the warm memoized bundle when one exists, and on a cold process it never constructs the embedder (a full model load for sentence-transformers) — the count comes from a store-only client and the geometry metadata from the prior vectorstore-index artifact. |
| pp_vectorstore_checkA | Diagnose vectorstore reachability + dimension (read-only; never writes). This tool is inherently a dry-run diagnostic: it NEVER constructs the store client and NEVER writes to disk. It runs the §6.5 conflict matrix with two bounded, fail-closed live probes wired in:
|
Prompts
Interactive templates invoked by user choice
| Name | Description |
|---|---|
No prompts | |
Resources
Contextual data attached and managed by the client
| Name | Description |
|---|---|
No resources | |
TDQS
Scored across 27 tools
Many tools have overlapping or unclear purposes, especially without descriptions. For example, pp_run_status, pp_init_run, pp_list_runs are ambiguous; pp_config_path, pp_config_show, pp_config_get, pp_config_set are confusingly similar. Even with partial descriptions, agents would struggle to select the correct tool.
All tools use the 'pp_' prefix and snake_case. Most follow a verb_noun pattern (e.g., init_project, list_sources, store_artifact). There are minor deviations like pp_run_status (noun_verb) and pp_next_phases (adjective_noun), but overall the naming is fairly consistent.
27 tools is on the heavier side but not excessive. The server covers multiple sub-domains (config, runs, phases, sources, artifacts, indexing). Some tools could be consolidated (e.g., merging config tools or status tools), but the count is manageable.
The tool set covers many lifecycle operations for runs, phases, sources, artifacts, and indexing. However, there are notable gaps: no delete operations for sources, artifacts, or runs; no update operations; and lack of tools for user management or permissions. The domain is partially covered.