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 joenandez/spectre --skill spectre-create_prgit clone --depth 1 https://github.com/joenandez/spectreWrote 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/joenandez/spectre/spectre-create_pr)<a href="https://agentmods.dev/skills/joenandez/spectre/spectre-create_pr"><img src="https://agentmods.dev/badge/skills/joenandez/spectre/spectre-create_pr/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/joenandez/spectre/spectre-create_pr"><img src="https://agentmods.dev/badge/skills/joenandez/spectre/spectre-create_pr.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector pass
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 | $0.00046 | $0.00994 |
| Opus 5.5 | $0.00018 | $0.00398 |
| Sonnet 5.5 | $0.00009 | $0.00199 |
| Haiku 4.5 | $0.00005 | $0.00099 |
Grade A, and why
spectre-create_pr 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 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.
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 — 52 lines — stays where its author put it; the contents beside it link to each section on GitHub.
create_pr
Produce a grounded draft PR.
Inputs
$ARGUMENTS:TARGET_BRANCH(defaultorigin/main), feedback focus, verification, all-or-noneEXPECTED_BASE_SHA/EXPECTED_HEAD_SHA/EXPECTED_DIFF_SHA256, or--orchestrated.- Orchestrated
--pr-phase pending|final-update: pending needs complete candidate tuple, local verificationRUNNING, and returns URL/body; final-update needs its URL/body, same tuple, andFINAL_VERIFICATION_SUMMARY, updating existing draft Testing only. A parent may supply a frozen work-ID set. Orchestrated Create PR does not write a work record; an unchanged candidate is a no-op. - Resolve just-in-time: branch (not
main/master), fetch target and derivePR_BASE,BASE_SHA,HEAD_SHA, canonicalgit-diff-v1DIFF_SHA256over binary/full-index/no-color/no-renames{BASE_SHA}...{HEAD_SHA}, commits, branch/commit issue ref,gh, unpushed commits. - Expected fields are all-or-none. After fetch compare live tuple and clean candidate worktree (committed review/proof artifacts allowed; ignored lifecycle excluded). Tracked or non-ignored untracked changes return
PR_CANDIDATE_STALEbefore push, create, or edit. Withoutgh, return title/body for manual draft.
Working Set
Target-to-HEAD diff and commit log, issue reference, and any GitHub PR template (required sections win).
Outputs + DONE
- Conventional
type(scope): summary: <70 chars, imperative, lowercase after colon, no period; grounded body scales to change. Standalone/pending opensgh pr create --draftonly without an open branch PR, else updates it.
DONE: fetched tuple is verified; every factual claim is grounded; Testing honestly reflects diff tests and verification, never turns advisory non-green into pass; sourced/placeholder Why; no secret/credential/PII; only a draft is opened or updated, never duplicated.
Method / guardrails
- Ground: What is behavior; Why is issue, commits, branch, or
<!-- WHY: motivation not found in commits/issue — fill in -->; visible How/trade-offs; Testing reports supplied verification, changed tests, or none. - Scale: trivial What/Why/Closes; standard Summary/Changes/Testing/Closes; complex adds trade-offs, breaking/rollback, UI/CLI evidence, reviewer focus. Derive type/scope from change; add found issue links only.
- Verify before side effects: map every claim to diff/commit/issue, drop unsupported claims/secrets, verify tuple, clean candidate, and this branch's own open PR (
gh pr list --head; never another branch's or closed), then push. - Work-record ownership: in orchestrated mode return PR evidence to Ship without writing a work record. In standalone mode only, when no Execute or Ship parent owns history, invoke
Skill(spectre-work-record)once at the terminal boundary after the draft exists; that write and explicit historical correction are unchanged; report capture failure/recovery input without blocking draft. Refresh a stale parent candidate; a draft is not merged. - Draft lifecycle: pending grounds
RUNNING, pushes, updates that open PR, else creates the draft, and returns its PR identity/URL against the frozen work-ID set for Ship to associate. Final-update rechecks its tuple/clean candidate; if repairs changed the tuple, refresh candidate-sensitive claims under freshness, grounding, secret gates, verify clean repaired HEAD, pushes, re-resolves/rechecks live tuple, thengh pr editonly Testing fromFINAL_VERIFICATION_SUMMARY; never mark ready.
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 Changed · -18 tokens per session d42e0042d9cd
- 1mo ago Changed 32902327e23f
- 1mo ago Changed 3f257e8174f3
- 1mo ago Changed · +8 lines 3fa2c1d4d56e
- 2mo ago Changed · -9 lines · -66 tokens per session f5054f0ec591
- 2mo ago First seen · 53 lines · 130 tokens per session scan A 5467becd0ee9
spectre-create_pr is a skill published in the GitHub repository joenandez/spectre (164 stars, last pushed 1mo ago), licensed MIT. It adds 46 tokens to every session and 994 once invoked, about $0.0002 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 2026-08-30.
spectre-create_pr compared with skills from other repositories
| Mod | Type | Stars | Last push | Licence | Scan | Tokens | Works with |
|---|---|---|---|---|---|---|---|
| spectre-create_prjoenandez/spectre | Skill | 164 | 1mo ago | MIT | A | 46 | Claude Code |
| merge-batchdiegosouzapw/OmniRoute | Skill | 83k | 6d ago | MIT | A | 46 | Claude Code |
| prowler-commitprowler-cloud/prowler | Skill | 15k | yesterday | Apache-2.0 | A | 44 | Claude Code, Codex |
| gh-auth-isolationgithub/gh-aw | Skill | 5.6k | yesterday | MIT | A | 20 | Claude Code, Codex |
| github-skillzeenie-ai/OpenCompany | Skill | 1.1k | yesterday | MIT | A | 51 | Claude Code |
| comet-githubrpamis/comet | Skill | 3.3k | yesterday | MIT | A | 72 | Claude Code, Codex |
| release-helperSahir619/fable-method | Skill | 2.3k | 1mo ago | MIT | A | 43 | Claude Code |