hermes-trading-research-agent-work-package: Instructions file for Codex

AGENTS.md

hermes-trading-research-agent-work-package AGENTS.md is an instructions file for Codex, OpenCode from tradermonty/hermes-trading-research-agent-work-package. It costs 4,704 tokens per session, scanned A, original, MIT.

A project guide for coding agents working on Hermes, a trading-research work package. It records the required work order, current version status, source of truth, and constraints.

In plain words
What is it for?
Following the project’s implementation phases, checking Hermes compatibility, maintaining scheduled jobs, and working with trade tickets, persistence, journals, and audit rules.
Why use it?
It gives agents one place to check what has already shipped and which instructions take priority, reducing incorrect or duplicated changes.

Instructions file for CodexOpenCode

Written for Codex and OpenCode: the file is AGENTS.md.

This is tradermonty/hermes-trading-research-agent-work-package's own configuration. It tells Codex and OpenCode how to work on hermes-trading-research-agent-work-package itself, so it is not a mod to install elsewhere. Copy it as a starting point and replace the rules that are about this project. Everything hermes-trading-research-agent-work-package configures →

Reuse

Borrowing it

Nothing to install: this file belongs to tradermonty/hermes-trading-research-agent-work-package. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.

Copy the file
curl -O https://raw.githubusercontent.com/tradermonty/hermes-trading-research-agent-work-package/main/AGENTS.md
Clone the repo
git clone --depth 1 https://github.com/tradermonty/hermes-trading-research-agent-work-package

Made for: Codex, OpenCode.

Wrote 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.

agentmods badge for hermes-trading-research-agent-work-package AGENTS.md

README.md
[![agentmods](https://agentmods.dev/badge/instructions/tradermonty/hermes-trading-research-agent-work-package/agents-md/github.svg)](https://agentmods.dev/instructions/tradermonty/hermes-trading-research-agent-work-package/agents-md)
Your own site
<a href="https://agentmods.dev/instructions/tradermonty/hermes-trading-research-agent-work-package/agents-md"><img src="https://agentmods.dev/badge/instructions/tradermonty/hermes-trading-research-agent-work-package/agents-md/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.

agentmods 80×15 button for hermes-trading-research-agent-work-package AGENTS.md

Your own site · 80×15
<a href="https://agentmods.dev/instructions/tradermonty/hermes-trading-research-agent-work-package/agents-md"><img src="https://agentmods.dev/badge/instructions/tradermonty/hermes-trading-research-agent-work-package/agents-md.svg" alt="Reviewed on agentmods" width="80" height="20"></a>
Per session 4,704 This file is loaded in full into every session.
When invoked 4,704 The same file — it is already loaded in full.
Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe.
Origin original No closer match found in the catalogue.
Token cost

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.

ModelPer sessionOnce invoked
Fable 5.1 $0.04704 $0.04704
Opus 5 $0.02352 $0.02352
Sonnet 5 $0.00941 $0.00941
Haiku 4.5 $0.00470 $0.00470

Measured 12d ago against content hash 24125b17b7e9, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-12, from the pricing page.

Security

Grade A, and why

hermes-trading-research-agent-work-package AGENTS.md scanned grade A with 0 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 12d 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.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

AGENTS.md · 198 lines

How it starts

The opening of the file, as written. The whole thing — 198 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Work Order for Coding Agents

Status (as of v0.1.6 — 2026-05-24)

This document was originally written as a pre-v0.1.0 coding handoff. The MVP (v0.1.0), CI / docs polish (v0.1.1), cron timezone single-source refactor (v0.1.2), the {{TIMEZONE}} prompt template + verified-Hermes-v0.14.0 docs round (v0.1.3), the generator-ownership + determinism round (v0.1.4 / TICKET-004a / B-2a), the trade-ticket primitive + Operational soak procedure + audit-field polish (v0.1.5 / TICKET-009 / B-3), and the trade-ticket persistence + journal bridge + upstream workflows drift guard + typo-guard polish (v0.1.6 / TICKET-010 / TICKET-004b / B-2b) have all shipped. Phase 2 is now ✓ done (B-2a + B-2b shipped). The bundle catalogue is 10 (/trade-ticket added in v0.1.5). The Phase / Ticket structure below is kept for historical context — it captures the original intent and the canonical source-of-truth ordering. For current state, this table is authoritative; the Phase bodies below are not.

Phase / Ticket Status Where it landed
Phase 0 — Verify Hermes capabilities ✓ done docs/03-hermes-compatibility-notes.md (verified table + operational findings, Hermes v0.14.0; v0.1.3 added verified entries for bundles reload, MCP config location, and cron toolset restrictions)
Phase 1 — External-linked skill integration ✓ done config.yaml:skills.external_dirs, scripts/validate_upstream_index.py, docs/04-skill-integration-strategy.md
Phase 2 — Bundle generation ✓ done (B-2a + B-2b shipped) Canonical index validation (scripts/validate_upstream_index.py) + generator ownership (x-generated: contract, write-if-changed, update_external_config idempotent, --force-overwrite escape hatch double-gated by REQUIRE_SYNC_WRITE=1 + REQUIRE_FORCE_OVERWRITE=1) + 7 determinism tests in tests/test_sync_determinism.py shipped as TICKET-004a / B-2a. B-2b (TICKET-004b) closes Phase 2: upstream workflows/*.yaml read-only drift guard for the 3 overlap workflows + 2 ignored-with-reason workflows; sync() write path unchanged (all 3 overlap bundles are x-generated: false). See tests/test_upstream_workflow_adapter.py for the 5 drift checks and docs/04 "Current bundle-composition SoT (as of B-2b)".
Phase 3 — Cron presets ✓ done data/schedule-presets.yaml (single source of truth for schedule / name / skills / prompt_file / timezone), cron/create_cron_jobs.py (host-TZ warning + {{TIMEZONE}} prompt template expansion from v0.1.3, priority shell env > .env > preset YAML > LA), cron/create_cron_jobs.sh wrapper, tests/test_schedule_drift.py, tests/test_prompt_template.py. cron run requires a job_id rather than a human name in Hermes v0.14.0 — documented in docs/03-hermes-compatibility-notes.md.
Phase 4 — Safety and policy gates ✓ done SOUL.md, tests/test_output_safety.py, tests/test_required_sections.py, tests/fixtures/sample_outputs/
Phase 5 — Optional vendored mode scoped (008a), impl deferred (008b) TICKET-008a documented the three-mode model + vendor-manifest.json shape + destructive-skeleton warning in docs/04. TICKET-008b (the actual non-destructive writer + manifest provenance + patches pipeline + skills drift detection) remains open.
Phase 6 — Release and documentation ✓ done v0.1.0 / v0.1.1 / v0.1.2 / v0.1.3 / v0.1.4 / v0.1.5 / v0.1.6 published on GitHub Releases. README + README.ja + CHANGELOG ship with each release.
TICKET-001..003, 005..007 ✓ done See v0.1.0 changelog.
TICKET-004a (generator ownership + determinism) ✓ done (shipped in v0.1.4) See the [0.1.4] section of CHANGELOG.md; docs/09-coding-tickets.md for the 004a/004b split rationale; tests/test_sync_determinism.py for the contract tests.
TICKET-004b / B-2b (upstream workflows adapter, closes Phase 2) ✓ done (shipped in v0.1.6) scripts/sync_claude_trading_skills.py adds 3 public symbols (UPSTREAM_OVERLAP_SLUGS, UPSTREAM_IGNORED_WORKFLOW_SLUGS, load_upstream_workflow); sync() write path unchanged. tests/test_upstream_workflow_adapter.py adds 5 named drift tests (13 parametrize cases): file-exists × 3, required-subset × 3, optional-in-mapping × 3, canonical_source marker × 3, inventory-classified × 1 (filesystem-side, also disjoint-checks the two slug sets). docs/04 carries SoT split + projection-vs-drift-check table (display_name / cadence are documented but not drift-checked; artifacts → required_outputs / steps / decision_gate / manual_review / when_to_run are not projected in v0.1.x — B-2c candidates) + inventory classification table (overlap: 3, ignored: 2 with reasons). Makefile sync-external-write comment fixed "nine → ten SKIP lines". With CLAUDE_TRADING_SKILLS_REPO: 162 → 175 passing (further +1 to 176 from the TICKET-010 polish fixture). Without: 162 passing, 14 skipped (13 new B-2b env-gated + 1 pre-existing B-2a env-gated case; the polish fixture is schema-only and runs in both modes). See docs/09-coding-tickets.md TICKET-004b entry.
TICKET-008a (vendored mode scope, docs) ✓ done (on main, folds into next release) docs/04 "Vendored mode (TICKET-008)" section + docs/09 008a/008b split. Documents the three-mode model and the destructive-skeleton warning. No code.
TICKET-008b (vendored mode implementation) open (deferred) Contract is the 008a docs. Surface: manifest schema + non-destructive re-vendor + patches pipeline + skills drift detection. See docs/09-coding-tickets.md TICKET-008.
TICKET-015 (upstream status inspection) ✓ done (on main, folds into next release) scripts/upstream_status.sh + make upstream-status (working-tree-safe pre-pull inspection: resolved ref / dirty / incoming / risk label). tests/test_upstream_status.py (presence + behavioral temp-git fixture). Docs in docs/03, docs/04 (three-layer table), docs/09, README EN/JA. Complements TICKET-004b drift guard (before-pull vs after-pull). Deferred: TICKET-015b read-only /upstream-status slash command.
TICKET-009 / B-3 (/trade-ticket bundle + schema + audit-field polish) ✓ done (shipped in v0.1.5) schemas/trade-ticket.schema.json (5-value status enum, approval.required: const true, APPROVED branch tightens numeric fields to non-null, approval.reviewer: minLength 1, created_at / decided_at carry both format: date-time and an explicit ISO-8601 pattern), skill-bundles/trade-ticket.yaml (x-generated: false, 5 operator verbs, positive boundary, confirmed.* re-type, mismatch handling), data/skill-mapping.yaml entry, escalation pointers in /swing-opportunity-daily + /pre-market-routine, tests/test_trade_ticket_schema.py (22 cases — validator built with format_checker=FORMAT_CHECKER; 9 schema negatives including blank-reviewer + invalid-timestamp) + 14 fixtures. Bundle count goes from 9 to 10. See docs/04-skill-integration-strategy.md "Trade ticket primitive" and docs/09-coding-tickets.md.
TICKET-010 (Trade ticket persistence + journal bridge, + typo-guard polish) ✓ done (shipped in v0.1.6) .gitignore adds tickets/ + *.ticket.yaml safety net. .env.EXAMPLE + distribution.yaml:env_requires declare HERMES_TRADE_TICKET_DIR (optional, default ${HOME}/trading-research/tickets). schemas/trade-ticket.schema.json adds optional journal_bridge object (target: const trader-memory-core, action: enum [register_thesis, update_thesis, postmortem], optional thesis_status / notes, additionalProperties: false). skill-bundles/trade-ticket.yaml instruction body gains save-path hint, journal-bridge handoff, silent-write positive form. tests/test_trade_ticket_schema.py +5 tests +2 parametrize rows (matrix 9 → 11 including the v0.1.6 typo-guard polish row); tests/test_package_structure.py +1 manifest declaration test. 3 new fixtures (14 → 17 with the polish). Suite 155 → 163 schema-only contribution (162 → 163 over the v0.1.5 baseline net of the polish). See docs/04-skill-integration-strategy.md "Ticket persistence and journal bridge" and docs/09-coding-tickets.md.

Read the full file on GitHub · 198 lines

Changes

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.

  1. 12d ago First seen · 198 lines · 4,704 tokens per session scan A 24125b17b7e9

Subscribe to this mod's changes

hermes-trading-research-agent-work-package AGENTS.md is an instructions file published in the GitHub repository tradermonty/hermes-trading-research-agent-work-package (38 stars, last pushed 3mo ago), licensed MIT. It adds 4,704 tokens to every session, about $0.0235 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.

Related

Other instructions, from other repositories

spec-kit AGENTS.md

AGENTS.md instructions for github/spec-kit, covering agents.md, about spec kit and specify, quickstart — add a new integration in 5 steps, integration architecture and integrationmanifest — file tracking.

github/spec-kit · 7,126 tokens

next.js AGENTS.md

AGENTS.md instructions for vercel/next.js, covering next.js development guide, codebase structure, monorepo overview, core package: packages/next and other important packages.

vercel/next.js · 7,296 tokens

codex AGENTS.md

AGENTS.md instructions for openai/codex, covering rust/codex-rs, the codex-core crate, code review rules, crate api surface and model visible context.

openai/codex · 5,153 tokens

vscode buildNext.instructions.md

Working notes and architecture documentation for the new esbuild-based build system in build/next. Use when making changes to the new build pipeline (transpile/bundle commands, NLS plugin, source-map handling, resource copying, or self-hosting watch tasks).

microsoft/vscode · 6,785 tokens

langchain AGENTS.md

AGENTS.md instructions for langchain-ai/langchain, covering global development guidelines for the langchain monorepo, corridor security analysis, project architecture and context, monorepo structure and development tools & commands.

langchain-ai/langchain · 4,469 tokens

vscode oss-third-party-notices.instructions.md

Instructions for microsoft/vscode, covering vs code oss third-party-notices pipeline, architecture, pipeline flow in ci, applying the notice (cutover) and fallback chain (never fail the build).

microsoft/vscode · 5,001 tokens