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 skills add RockyHong/super-bootstrap --skill loggit clone --depth 1 https://github.com/RockyHong/super-bootstrapWrote 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/skills/rockyhong/super-bootstrap/log)<a href="https://agentmods.dev/skills/rockyhong/super-bootstrap/log"><img src="https://agentmods.dev/badge/skills/rockyhong/super-bootstrap/log/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/skills/rockyhong/super-bootstrap/log"><img src="https://agentmods.dev/badge/skills/rockyhong/super-bootstrap/log.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 | $0.00152 | $0.02085 |
| Opus 5.5 | $0.00061 | $0.00834 |
| Sonnet 5.5 | $0.00030 | $0.00417 |
| Haiku 4.5 | $0.00015 | $0.00209 |
Grade A, and why
log 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 17d 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.
How it starts
The opening of the file, as written. The whole thing — 41 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Log — Capture Front Door
Muscle-memory capture, run gateway-inline in the calling session — no dispatch. Takes one observation or a batch, classifies each into BUG / DEBT / GAP, checks against open cards, and writes each clear entry as docs/work/{ID}.md in docs/work/TEMPLATE.md's shape. Capture is unconditional about worth (bugs, debt, design gaps, unverified ideas all land); the real/worth/now call runs at /super-bootstrap:triage on pickup.
All new cards route through this door — user-initiated and Claude-initiated captures alike. Hand-copying docs/work/TEMPLATE.md with an ID from step 4's generator is the same door, by hand.
When it fires
- User — explicit
/super-bootstrap:log <observation>, or natural-language "log this / track that / note this down". Bare/super-bootstrap:log→ ask what to capture (one line). - Claude — its own captures: a bug spotted mid-task, out-of-scope findings a review or returned subagent surfaced. Batch them into one pass.
A card is owed only for deferred or dropped work — work that exits the current flow incomplete. Work completed in-flow carries no card debt. The trigger is completion-state (observable), not worth.
Out of lane: root-cause investigation (the triage lane), card deletion (the resolving session), rewriting existing blocks (threads are append-only). Feature ideas and unverified hunches are in lane — they log as GAP.
Procedure
- Read
docs/work/README.md— categories and thread contract. README absent → write nothing; route: "run/super-bootstrap:harness-bootstrapto plant the work substrate, then re-log." When present: runbash <skill-base>/assets/closed-forks.shfrom the repo root, issued as its own tool call —<skill-base>is theBase directory for this skill:path surfaced at invocation. It resolvesdocs/decisions.md§ Closed Forks to one line per row, or toabsent, and ends with the outcome checklist step 2 emits. A non-zero exit names the file and the cause → surface the script's message and hold for the user's pick: they fix the doc and re-log, or the batch logs on withdecisions.md § Closed Forks — unresolvedas each entry's step-2 outcome. The doc fix is the user's. Readdocs/parked.md's header anddocs/outward/README.md's header too — each skipped silently when absent. - Classify each entry — BUG (broken behavior) / DEBT (works but rotting) / GAP (design hole or unverified idea). Three gates per entry, in order:
- Mover — ask once, whose hands move the next step — the repo's, yours, or an outside party's? The repo → a card, through the gates below. Yours or an outside party's (a taste sitting, a line-by-line review, a portal registration, a reply to wait for) → an outward thread at
docs/outward/{ID}.mdwhen the folder exists, itsOUTID from step 4: origin block only,docs/outward/TEMPLATE.md's shape (H1# {ID} — {summary}, thenLogged:/Source:/Next move:/Waiting on:/Repo tail — fires on:, plusOwning card:when a card waits on that step; drop the template's leading comment). Ask for the repo tail in one line whenever the observation doesn't name one — the answer lands the thread; only an item with no tail at all drops, per the container README. Nodocs/outward/but a flatdocs/outward.md→ the container predates the folder shape: write nothing for that entry and route: "run/super-bootstrap:harness-bootstrap— its sync splitsdocs/outward.mdintodocs/outward/— then re-log." Neither present → a card whoseProblem:line nameswaiting on {party}, which the board's wait override reads. Mixed → split at capture: the repo's remainder is the card, the other step the thread, carryingOwning card:when the card waits on that step. - Card shape — every card names something to do, fix, or decide. A standing-watch entry (monitor / revisit later) rephrases as an action; names a trigger +
docs/parked.mdexists → park as a### {ID} — {summary}entry, itsPARKID from step 4; otherwise return it with the real fork: wire an observer at the fire-moment, or drop. - Closed forks — hold each entry against the rows step 1's script printed, and emit one outcome line per entry that reaches this gate, in the reply, whatever the read found:
decisions.md § Closed Forks — collides: <row>/decisions.md § Closed Forks — no collision/decisions.md § Closed Forks — absent(or— unresolved, per step 1's pick). Collides → surface the fork (open the row for its Because + Ref) instead of logging; genuinely new counter-evidence → ask as a reopen question. Never silently re-log a closed direction.
- Mover — ask once, whose hands move the next step — the repo's, yours, or an outside party's? The repo → a card, through the gates below. Yours or an outside party's (a taste sitting, a line-by-line review, a portal registration, a reply to wait for) → an outward thread at
- Dedup — the call is the user's, not capture's. Grep the open cards — plus the
docs/outward/OUT-*.mdthreads when present — for each entry's subject (title keywords,Area:paths for cards; H1 summary orRepo tailpath for outward threads); open a suspected match to confirm coverage. A suspected dup — pure duplicate or duplicate-with-new-fact alike — writes nothing yet: surface the entry, the covering card or thread (ID + the covering line), and the pick — amend (append the new fact) / new card (genuinely distinct) / drop. Execute the pick: amend → append## Amendment — {date} · {source}at the owning card's or outward thread's end, the new fact alone, nothing restated — on an outward thread whoseNext move/Waiting onthe fact changes, that block restates those two lines and leads from there. - Assign IDs — run
bash <skill-base>/assets/new-id.sh <CAT>...from the repo root, once for the whole batch, issued as its own tool call: one category argument (BUG/DEBT/GAP/OUT/PARK) per file about to be written (a mixed entry = card +OUT), in entry order. It prints one ID per argument, in the same order —{CAT}-xxxx, a random 4-char suffix the script redraws on any ID already in the repo or its commit messages. Use each printed ID verbatim for its entry's file name and heading. Never compose, edit, or increment an ID by hand, and never derive one from the live files or another ID — a hand-made ID is the cross-clone collision the generator exists to prevent. A non-zero exit → surface its message and write nothing. Legacy numeric IDs (GAP-101) on existing cards stay valid; there is no counter to bump. - Write each clear entry as
docs/work/{ID}.md— origin block only,TEMPLATE.md's shape (H1 +Logged:/Source:/Problem:/Area:/Prior:; drop the template's leading comment). Stamp today's date. Capture the claim faithfully — this is the richest-context moment; the pickup session reads it cold.Prior:is a one-line suspected cause at most — later work falsifies it. Scale-module fact fields ride the origin block as further field lines when known at capture. - An entry too thin to classify or phrase as a readable card → hold it unwritten and ask its one discriminating question (e.g. "broken now, or works-but-suboptimal?"); the answer lands it.
What ships with it
2 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
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.
- 17d ago Changed 4e5bd7a5926e
- 1mo ago Changed · -12 tokens per session 9176921bfc6c
- 1mo ago First seen · 41 lines · 164 tokens per session scan A 57cf81f5cc29
log is a skill published in the GitHub repository RockyHong/super-bootstrap (3 stars, last pushed today), licensed MIT. It adds 152 tokens to every session and 2,085 once invoked, about $0.0006 per session on Opus 5.5. 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 2026-08-31.
Other skills, from other repositories
plan-to-tasks
Transform a plancreation planv{N}.md into a structured tasks.yaml + per-WP workpackages/.yaml using parallel Sonnet sub-agents. Use after plan-creation finishes when you need execution-friendly task structure.
traction-eos
Implement the Entrepreneurial Operating System (EOS) to align vision and execution across a company. Use when the user mentions "EOS", "Entrepreneurial Operating System", "V/TO", "quarterly rocks", "Level 10 meetings", "accountability chart", "IDS process", "my company feels chaotic", "we keep having the same…
great_cto
Use when the CTO describes a feature, task, or project goal. Orchestrates the full SDLC pipeline automatically based on project type.
hermit-scribe
File a GitHub issue or post a comment on an existing issue via a GitHub App bot identity. Use when the operator says "file as a GH issue", "open an issue for PROP-NNN", "report this to the tracker", "add a comment to issue.
stale-proposals
Audit this repository's local proposal queue for stale entries: proposals still marked proposed, deferred, or accepted even though the work already shipped, plus old proposals whose subject has become inert. Reconcile proposals against plugin changelogs and first-parent Git history, automatically resolve only…
atlas-present
Generate a polished HTML presentation page and Obsidian Canvas for big releases — new products, takeovers, major migrations. Non-technical audience. Use when asked to "present this", "release announcement", "show what we built", or "stakeholder update".