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.
git clone --depth 1 https://github.com/assafkip/kipi-systemWrote 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/agents/assafkip/kipi-system/step-orchestrator)<a href="https://agentmods.dev/agents/assafkip/kipi-system/step-orchestrator"><img src="https://agentmods.dev/badge/agents/assafkip/kipi-system/step-orchestrator/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/agents/assafkip/kipi-system/step-orchestrator"><img src="https://agentmods.dev/badge/agents/assafkip/kipi-system/step-orchestrator.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 | $0.00037 | $0.04065 |
| Opus 5.5 | $0.00015 | $0.01626 |
| Sonnet 5.5 | $0.00007 | $0.00813 |
| Haiku 4.5 | $0.00004 | $0.00407 |
Grade C, and why
morning-orchestrator scanned grade C with 1 finding 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 1mo 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.
Recursive force deletehighDestructive command
rm -rf with a variable or a broad path is one typo away from removing the wrong tree.
find q-system/.q-system/agent-pipeline/bus/ -maxdepth 1 -type d -mtime +3 -exec rm -rf {} \; How it starts
The opening of the file, as written. The whole thing — 239 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Agent Pipeline Orchestrator
This step replaces the monolithic morning routine with a phased agent pipeline. Each phase spawns sub-agents via the Agent tool. Agents communicate through JSON files in the bus/ directory.
MANDATORY READS (before Phase 0 -- do not skip)
You are about to execute a 9-phase agent pipeline. Before spawning ANY agent, read these files. If you feel like you already know what to do from the skill prompt -- you don't. Every session that skipped these reads re-discovered known bugs and broke guardrails.
Read 1: Preflight -- Tool Manifest + Known Issues
Read: q-system/.q-system/preflight.md (offset: 1, limit: 136)
Sections 1-2: What tools work, what's broken, fallback chains, and 10 known issues (KI-1 through KI-10) that have burned real sessions. The preflight agent tests connections, but it doesn't know about KI-5 (LinkedIn search returns vendors) or KI-6 (generic subs produce 0 leads) or KI-9 (new positioning). You do.
Read 2: Preflight -- Execution Gates + Action Cards
Read: q-system/.q-system/preflight.md (offset: 246, limit: 75)
Sections 5-6: When to halt, how to log steps, and the rule that "card delivered" is NOT "done" -- only founder-confirmed actions update state files.
Read 3: Current Positioning
Read: q-system/canonical/decisions.md (limit: 60)
So agents generate content with current positioning, not deprecated messaging.
DO NOT proceed to Setup until all 3 reads are done. DO NOT rely on the SKILL.md prompt for execution details. This file is your authority.
Setup
DATE=$(date +%Y-%m-%d)
BUS_DIR="q-system/.q-system/agent-pipeline/bus/${DATE}"
AGENTS_DIR="q-system/.q-system/agent-pipeline/agents"
mkdir -p "${BUS_DIR}"
Context Management
Your context window will compact automatically at ~50% capacity (set by session-context.sh). After compaction, the PostCompact hook re-injects: operating mode, open loops, morning pipeline phase progress (which bus files exist), canonical positioning snapshot, and voice reminders. Do not skip steps or phases to save context. If context runs low, tell the founder what phase you're in and what remains. Never self-authorize skipping.
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.
- 1mo ago First seen · 239 lines · 37 tokens per session scan C 3e8d731596e6
morning-orchestrator is an agent published in the GitHub repository assafkip/kipi-system (121 stars, last pushed 2d ago), licensed MIT. It adds 37 tokens to every session and 4,065 once invoked, about $0.0001 per session on Opus 5.5. A static security scan graded it C with 1 finding (recursive force delete). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-09-03.
Other agents, from other repositories
compression-worker
Haiku-based agent for compressing verbose memories into concise summaries.
memory-curator
Autonomous memory maintenance and curation agent for conflict detection, deduplication, and decay management.
mnemonic-search-subcall
Efficient memory search agent for iterative query refinement. Executes targeted ripgrep searches and returns structured findings.
ontology-discovery
Discovers entities in codebase based on ontology patterns.
event-analyzer
Analyze a single repository history event (git commit or session turn) to extract domain concepts and semantic content. Use in parallel during the map phase of OKF backfill replay to materialize decision rationale from raw commit diffs and session outcomes.
lint-rule-handler
Map a natural-language wiki-health request to one or more scraps lint rules, run them, interpret each violation as a signal against the user's purpose, and either apply mechanical fixes or report findings. Use this agent for purpose-driven Scraps lint work.