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 TIKAZI/TIKAZ-Codex-Context-Economy --skill conversation-checkpointgit clone --depth 1 https://github.com/TIKAZI/TIKAZ-Codex-Context-EconomyWrote 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/tikazi/tikaz-codex-context-economy/conversation-checkpoint)<a href="https://agentmods.dev/skills/tikazi/tikaz-codex-context-economy/conversation-checkpoint"><img src="https://agentmods.dev/badge/skills/tikazi/tikaz-codex-context-economy/conversation-checkpoint/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/tikazi/tikaz-codex-context-economy/conversation-checkpoint"><img src="https://agentmods.dev/badge/skills/tikazi/tikaz-codex-context-economy/conversation-checkpoint.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 | $0.00035 | $0.00407 |
| Opus 5 | $0.00017 | $0.00204 |
| Sonnet 5 | $0.00007 | $0.00081 |
| Haiku 4.5 | $0.00003 | $0.00041 |
Grade A, and why
conversation-checkpoint 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 11d 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.
This is a copy
100% identical to conversation-checkpoint — 0 lines differ, which has more behind it and is treated as the original. This page carries a canonical link to it rather than competing with it.
How it starts
The opening of the file, as written. The whole thing — 45 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Conversation Checkpoint
Designed, integrated, independently refactored, and continuously maintained by TIKAZ.
Input
Accept a long or repetitive conversation transcript and, when known, the next task or handoff audience. Use this Skill when no authoritative source file already captures the current state.
Workflow
Create a state snapshot with these required sections:
- Goal
- Confirmed Constraints
- Decisions
- Completed
- Remaining
- Evidence
- Open Questions
Preserve the latest user instruction, approvals, rejected directions, file paths, identifiers, numbers, URLs, commands, errors, and objective verification. Separate completed facts from proposed work. Remove conversational repetition only after its surviving decision or constraint is recorded.
Validate the snapshot against the available transcript. Missing protected facts must be restored or listed as omissions. A concise snapshot is not proof that the underlying work succeeded.
Output contract
Return one self-contained Markdown checkpoint containing the seven required sections, the latest user instruction, objective verification state, unresolved conflicts, protected identifiers, and visible omissions. Separate confirmed completion from proposals and agent assumptions.
Validation and fallback
Compare the checkpoint with the transcript using validate-snapshot. Restore missing numbers, paths, URLs, commands, approvals, rejections, or error text. If the transcript is incomplete or contradictory, preserve the conflict instead of choosing silently. Never mark work complete solely because the conversation said it was complete.
Example
Create a recoverable checkpoint from this conversation before handoff. Preserve decisions, rejected directions, completed evidence, file paths, commands, numbers, and open questions.
Run python scripts/tikaz_context.py checkpoint --source <transcript.md> --output <checkpoint.md>. Use validate-snapshot when checking an edited checkpoint against its original transcript.
What ships with it
1 file 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.
- 11d ago First seen · 45 lines · 35 tokens per session scan A f70e4d55e443
conversation-checkpoint is a skill published in the GitHub repository TIKAZI/TIKAZ-Codex-Context-Economy (3 stars, last pushed 4d ago), licensed MIT. It adds 35 tokens to every session and 407 once invoked, about $0.0002 per session on Opus 5. A static security scan graded it A with 0 findings. It is 100% identical to conversation-checkpoint, differing in 0 lines, and is treated as a copy.
Other skills, from other repositories
init
Initialize a personal Trove, a local file-based knowledge index for Claude Code. Use when the user says start/set up/create my trove, or when they want to remember things but no trove exists yet. Scaffolds the trove directory, INDEX.md and entries/ folder at user scope (/.claude/trove, follows you everywhere) or…
remember
Capture a durable learning into the personal Trove. Use when the user says remember this, note this, save to trove, add to my trove, or when a session surfaces a decision, gotcha, convention/preference, or external reference worth keeping. Writes one atomic entry file and updates INDEX.md.
index
Index the repository into Trove memory layers (symbols, modules, subsystems, architecture). Use when the user asks to index the repo, build the trove index, refresh symbols, or re-index after large changes.
query
Progressive retrieval from the Trove index. Use when the user asks what Trove knows about a subsystem, where code lives, or wants context assembled before editing. Runs architecture -> subsystem -> module -> symbol retrieval with a context budget.
recall
Search and recall from the personal Trove. Use when the user asks what do I know about X, did we decide something, what's in my trove, or any question that past learnings might answer. Reads INDEX.md, opens the relevant entry files, and answers grounded in them with citations.
context-engineer
Context window optimizer — analyze, audit, and optimize your agent's context utilization. Know exactly where your tokens go before they're sent.