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.
npx agentmods add commands/hirohyun/grounded-copy/groundedgit clone --depth 1 https://github.com/HiroHyun/grounded-copyWrote 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/commands/hirohyun/grounded-copy/grounded)<a href="https://agentmods.dev/commands/hirohyun/grounded-copy/grounded"><img src="https://agentmods.dev/badge/commands/hirohyun/grounded-copy/grounded.svg" alt="Measured on agentmods" 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 | $0.00007 | $0.00565 |
| Opus 5 | $0.00003 | $0.00282 |
| Sonnet 5 | $0.00001 | $0.00113 |
| Haiku 4.5 | $0.00001 | $0.00056 |
Grade A, and why
grounded scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 3d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
What it actually says
Run exactly one of the four command lines below, then report what it prints
and change no other files. Pick the line by matching the argument you were
given against chat, copy, and off. An empty argument, or any other
value, takes the status line.
chat:
sh '${CLAUDE_PLUGIN_ROOT}/hooks/run.sh' grounded_tracker.py --set chat
copy:
sh '${CLAUDE_PLUGIN_ROOT}/hooks/run.sh' grounded_tracker.py --set copy
off:
sh '${CLAUDE_PLUGIN_ROOT}/hooks/run.sh' grounded_tracker.py --set off
Empty argument, or a value outside those three:
sh '${CLAUDE_PLUGIN_ROOT}/hooks/run.sh' grounded_tracker.py --status
Run the chosen line as written. This file interpolates no argument, so every command line in it is a constant; composing a new one would put a user-typed value on a shell command line.
For an unrecognized value, report the value that was rejected and the profile the status line names. The preference file holds still.
Exit codes on --set: 0 recorded, 1 persistence failure, 2 rejected value.
--status exits 0 in every state.
A recorded switch prints a governing directive after the status line. That directive supersedes every grounded-copy policy statement earlier in the transcript and governs from that turn onward.
If a command line carries a literal dollar-brace placeholder where an absolute
path belongs, the substitution failed: report that and stop. On a Windows
setup where sh is absent from PATH, run the same script through
hooks/run.cmd with double quotes around the path, matching the launcher swap
in skills/grounded-copy/references/setup.md; cmd reads ' as an ordinary
character.
Profiles: chat carries the core rules, copy adds the marketing register and
two closures, and off stops the session policy and the turn reminder. The
recorded value persists across restarts until another --set replaces it.
### Profile lifecycle in skills/grounded-copy/references/setup.md is the
one description of how the profile preference, the session policy, the turn
reminder, and the effective policy relate. The preference lives at
<config-dir>/grounded-copy/profile, where config-dir is $CLAUDE_CONFIG_DIR
when set and ~/.claude otherwise.
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.
- 3d ago First seen · 66 lines · 7 tokens per session scan A 22ad768ac518
grounded is a command published in the GitHub repository HiroHyun/grounded-copy (17 stars, last pushed 1mo ago), licensed MIT. It adds 7 tokens to every session and 565 once invoked, about $0.0000 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.
Other commands, from other repositories
churn-prevention
Win/loss interview + churn analysis + retention playbook. Focuses on retention (not acquisition) — where 80% of the current plugin is focused. For B2B SaaS with churn > 5%/month, consulting with renewal rate < 70%, or a recurring business on a plateau.
audit
Offer diagnostic via the Value Equation — scores 1-10 on each vector (Dream Outcome, Probability, Time Delay, Effort), pinpoints the critical bottleneck, proposes the top 3 concrete levers, and rewrites the offer in 1 paragraph. Recommended prerequisite before LP and script.
Email sequence (cold, warm, nurture, re-engagement). Generates 5-7 sequential emails with hook + agitation + proof + CTA, timing between touches, and context per touch. Tuned for B2B cold outreach — direct voice, no superlatives, specific proof.
help
Interactive decision matrix. Answer 3 questions and get the right command recommended. Useful for onboarding a new client or when you don't know where to start a session. Covers all 17 plugin commands.
client-onboarding
The client's first 30 days after the sale. A structured touch cadence, early quick wins to validate the purchase decision, visible value milestones. Cuts early churn (< 30 days) by a typical 40-60%. For consulting, agencies, cohort programs, mid-market B2B SaaS.
content-hub
30-90 day organic content roadmap. Topic × format × funnel stage × CTA. Healthy mix (60% educational, 25% entertainment, 15% promotional). For the founder/consultant who wants organic compounding over 6-18 months, not one viral post.