Bumblebiber/hmem-session-start

hmem-session-start

TRIGGER

Run at the beginning of any session where a Cortex project is active.

STEP 1: Activate project

Call load_project with the working project ID:

load_project(id: "P00XX")

Replace P00XX with the actual project ID (e.g., P0048).

load_project returns the project brief, recent O-Entry summaries, rules, and lessons. Do NOT call read_memory separately. load_project is the only activation action.

STEP 2: Noise Check

Do this immediately after load_project, before any other work.

Scan the output for:

• >4k tokens → invoke hmem-curate on this project first, then continue
• ✓ DONE items in Roadmap → update_memory(id, { irrelevant: true })
• Decommissioned / concept entries in Infrastructure → update_memory(id, { irrelevant: true })
• Old status snapshots in Overview (superseded by newer) → mark obsolete
• [-] prefix sections (e.g. P00XX.10 [-] Bugs) — orphaned schema artifacts → update_memory(id, { irrelevant: true }) for each
• Duplicate L2 sections (two nodes with same name) → mark the higher-numbered one irrelevant

Fix all of the above immediately. Do not note and defer.

STEP 3: Calibrate explanation depth

Read H0003 (IT Skills) — the scale is 1–9:

• 7–9 = Expert: use technical language directly, no padding, no basics
• 4–6 = Proficient: explain concepts but skip fundamentals
• 1–3 = Basics: explain with examples

Apply this calibration for the entire session. When explaining something in a domain, check the matching H0003 skill first.

STEP 4: O-Entry routing check

This step is critical. Every load_project call changes which O-entry receives session exchanges. If you called load_project on any project other than your working project — even briefly, even for administrative reasons (reconcile, curation, migration) — those exchanges were misrouted to the wrong O-entry.

After activating, check for misrouted exchanges:

read_memory(id: "O00XX")   ← the working project's O-entry (same seq as P00XX)

Look at the most recent batch. If exchanges are missing that you know happened (e.g. earlier in this session), they landed in another project's O-entry.

To find them: check O-entries for any other project you called load_project on during this session. Look for nodes created today with content matching your session.

To fix:

move_nodes(node_ids: ["O00YY.Z"], target_o_id: "O00XX")

Move the misrouted session/batch node to the correct O-entry.

Rule: Never call load_project on a secondary project without immediately re-calling it on your working project. Routing follows the last load_project call — always return control explicitly.

What gets injected automatically (first message)

The UserPromptSubmit hook injects the following into every session start:

• H-entries — top 10 by access count (ID + title)
• Active device apps — Apps list of the current I-entry (if device is set)
• Infrastructure favorites — any I-entry with favorite: true (e.g. reMarkable, shared server). Mark with update_memory(id="I00XX", favorite=true).
• Recent projects — 5 most recently updated P-entries

OUTPUT

After both steps, output exactly:

[CORTEX READY] Project: <name from load_project> Context loaded. Ready. [/CORTEX READY]

Then wait for the user's first message.